Web 2.0 tools and context
This chapter describes some of the key tools within the Web 2.0 suite which have potential corporate applications. These tools are essentially knowledge tools: they support knowledge creation, interaction and collaboration, networking and sharing. One can use them to mediate interaction between the personnel within an organisation, with the organisation’s customers or with business partners such as suppliers and vendors. This is not the intended function of these tools, which have their genesis in the role they play in the lives of Web users. It so happens that they have characteristics that make them useful for the management of working knowledge, knowledge which is a key productive resource for the delivery of industrial production or services.
We shall discuss a number of tools: some provide functionality which is used directly by users, such as wikis, blogs, social networking and RSS readers. Others provide underlying capability to the end-user tools, improving the usability of those tools: AJAX, REST, the semantic web. The tools we discuss offer a complete and mutually complementary suite for the management of knowledge in the corporate environment.
This is a key point of this chapter: that the Web 2.0 tools, although individually useful, should also be seen as a set of configurable components which, when working together, provide strong functional support for the production and exploitation of organisational knowledge. This platform of tools provides open-ended, highly flexible support for knowledge transformation activities and, I believe, wherever possible should be left to the devices of self-organising, dynamic groups and individuals to design, use and combine according to their requirements and capabilities. This is not only an emancipated and ethically correct approach, but I suspect that this is also the most likely way to achieve success.
The expression ‘Web 2.0′ emerged as part of a conference workshop conducted by O’Reilly media and was coined by Tim O’Reilly and Dale Dougherty as they discussed the aftermath of the 2001 dot com collapse.1 It was mooted that far from being over, the potential impact of the Internet was as great as ever but taking a different form. This form was exemplified by moves to new types of website, product or service, for example:
from file downloads on mp3.com to peer file-sharing using Napster;
These shifts contained some underlying principles which at least warranted the claim that a qualitative change was taking place. The ‘meme map’ in Figure 2.1 shows the major elements in this transition,2 but in essence the change was from a ‘Web 1.0′ which was predicated on the provision of data or services via a ‘server’ to a ‘client’, to a ‘Web 2.0′ which provided a platform for users to participate on an equal footing. A platform is a toolbox of capabilities which are open-ended, adaptable to many purposes and contain low levels of inbuilt constraint. Software tools and websites were appearing and constituting a Web toolbox platform which enabled any non-technical, low-budget user of the Internet to create, use and control information and information exchange as they chose. This context of user-driven, user-managed Web participation led to a rise in the provision of services (rather than finished products), which users could configure and mix, and enabled users to themselves provide ‘services’, even if that service was a stream of their own opinions in a blog.
Figure 2.1 Meme map – the gravitational core Source: http://oreilly.com/web2/archive/what-is-web-20.html
These services could start cheap, small and simple, but were infinitely scalable in both volume and function. One did not have to invest big to have an impact and it became possible to start a long Internet journey with a single step: the functions of the website could be upgraded in increments as a ‘perpetual beta’ (or stream of prototypes) without bigbang risks and costs. The speed and simplicity of communications products led to the development of interactive conversations (as opposed to just shouting into the void), allowing participation in persistent, prolonged conversations around shared areas of interest. Solid, dependable information and other forms of digital products, such as software, bookmarks, recommendations and reference works, emerged, based upon the collective intelligence of many participants. These outcomes became a service to others.
These tools were not invented to manage knowledge and information exchange in corporations. But they are tools which are integral to the system of knowledge creation and sharing which now dominates the Internet. They emerged because they play a role in capturing, exchanging, storing and classifying information symbols in a form which is persistent and shared by users across the physical and temporal ranges reached by the World Wide Web.
So is the expression convincing and durable, and are there essential characteristics in the definition of Web 2.0? There are certainly some tools in the Web 2.0 suite which existed before the expression appeared and even before one could imagine talking about Web 2.0. Other tools considered to be in the Web 2.0 set, such as Twitter’s microblogging technology, have emerged since the term was coined. And one needs to be ever mindful of sales hype in a rapidly moving marketplace of ideas in which the participants are constantly seeking points of differentiation in search of profit or acclaim. But there does seem to be a use for the concept, which embodies a new paradigm of interactivity, collaboration and self-service. This piece of language can be used to convey a message to corporate decision-makers, marketing personnel, designers of websites and e-business marketplaces, namely that forms of information exchange and generation are changing and that a product catalogue with secure credit card transactions is not enough to make a successful website, indeed that control over the future use of that information will restrict its generative power and the return on its investment.
The expression Web 2.0 provides an umbrella term which can be used to strategically harness the existing and rapidly multiplying tools into coherent yet responsive organisational strategies for interactions with customers, partners and between people within the corporate firewall. The elasticity and usefulness of the term are reflected in the emergence of terms such as ‘Library 2.0′, ‘Enterprise 2.0′ and ‘Government 2.0′ to reflect the same attributes of networked, interactive, accessible and participative information creation.3 Charles Leadbeater even talks of ‘Art 2.0′, in which the artistic avant-garde of the twenty-first century develop mechanisms and a culture which encourage people to create common works across the Internet.4 But we need to bear in mind, that the tools are a necessary, but not sufficient, part of the definition of Web 2.0. The way the tools are used is also definitive: using a wiki as a corporate intranet content management system, in which content is only allowed to be added by a certified administrator, is not Web 2.0.
A blog is simply a sequential log of writings or, more recently, video expression, published in reverse chronological order, such that the newest entries are at the beginning of the log. ‘Blog’ is an abbreviation of ‘weblog’, a form of personal record or narrative that first appeared on the Internet in 1997. It allows personal publishing with no editorial intervention or review. Blogging software became available in 1999 and since then has exploded as a form of expression and personal website management. The ‘blogosphere’, or collective community of all blogs, contains anything up to 180 million blogs in the English-speaking world alone (and over 70 million in Chinese). Almost all mainstream media publishers use journalist blogs and the most popular ones are overwhelmingly theirs.5
Blogs are generally personal in that they are owned by an individual (who may be representing a firm) and the blog is a vehicle for the expression of their views. But a blog can also be a ‘normal’ high-function website containing multimedia content with sophisticated presentation capabilities. A blog entry can be commented upon by others and responses can be built upon to form a cascading chain of comment and response. Each of these blog entries has a ‘permalink’ which is an individual web address, simplifying access to and distribution of a particular entry. For example, you may have read an insightful and particular comment in a blog: so you simply copy and send the permalink URL to a friend who may be interested. Blogs are often personal or political opinions, commentaries, news items and corporate announcements. They can become open forums for discussion and collaboration, but they still retain personal ownership. Blogs also generally provide subscriptions, creating RSS syndication feeds which can be subscribed to by others who will be notified when a new comment is made or responded to. ‘NightJack’ for example is the depressing but compelling blog of an anonymous policeman who won the Orwell prize for journalism. Unfortunately, NightJack (see Figure 2.2) was ‘outed’ in 2009 and forced by his employer to cease blogging.
Figure 2.2 The NightJack blog (now unavailable) Source: http://nightjack.wordpress.com
There are several ways to obtain and use a blog. There are hosting services, such as Bloglines or blogger.com, which allow one to set up a personal blog on the Internet and use the various blog functions for free. One can download free blogging software to run on a web server, a typical example being Wordpress, which uses the open source MySQL database management software. Or there are blog functions existing within corporate products such as Microsoft Office Sharepoint Services, the Confluence or Socialtext wiki products, or as an extension in MediaWiki.
The blog is a powerful broadcast mechanism with many corporate applications. A personal blog can be used to communicate with co-workers, a collaborative blog to communicate between teams (for example at shift change) and a corporate blog can be used to broadcast newsworthy events to the enterprise.6 As an outward-facing medium, it can be used by nominated business specialists to post information about products and services, plans and directions, and general chit-chat. The first famous corporate blogger was Robert Scoble (see Figure 2.3), who gave Microsoft a human and sometimes critical face; he now blogs independently of that company. In a recent interview he noted that corporate blogs seem lifeless because they are like press releases, ‘new, friendly, cuddly press releases’.7 Authenticity, the representation of a true, personally held set of beliefs, seems an important component to a successful blog.
Figure 2.3 Robert Scoble’s blog Source: http://scobleizer.com/
In organisations whose core business is information, such as entertainment, news media or lifestyle programming, blogs take on an operational hue: their production is the product.8 But Blogs are not just a product for the media; they also support other journalistic processes. In one survey, 60 per cent of journalists use them for research, 51 per cent to create opinions and 46 per cent to discover themes which interest the public.9 Further research into the impact of blogs on product decisions demonstrated that although 8 per cent of Internet users use blogs for information, of these, 54 per cent have built an opinion on a blog entry.10 Of course, as Dan Gilmour points out, these tools and more are available to everyone, not just to professional journalists.11
Blogs can be used in the monitoring and management of issues and emerging problems. Koller and Alpar analysed the use of Internet blogs in issue management in the public sphere: they cite two examples of issue conflagration through blogs. The first cost the Kryptonite bicycle lock company over $10,000,000 and considerable reputation when an expensive lock was shown in an uploaded video to be opened using a Bic pen.12 The second example was the ringtone vendor Jamba, whose staff attempted to defend their pricing model against a critical, but amusing, blog entry which was becoming popular with young people. When this was discovered, it led to a ‘blogstorm’ and the original critical blog appearing (because of the increase in the number of links to that offending blog) on the first page of search engines. The loss of reputation to Jamba was significant. The bottom line for companies is that although cases like these are rare, and the cost of blog monitoring can be relatively high, certain industries (particularly automobile, retail, telecommunications and transport) should pay attention to blogs and their criticisms (34 per cent of firm-specific blog entries are negative).
Although numbers are perhaps starting to plateau, Internet blogs are now an established part of much personal and professional information sharing and mainstream product information management. Blogging will become increasingly important, although its form will change as new modes (such as real-time microblogging) become available.13 If only for defensive purposes, firms need to monitor Internet blogging activity. In a time where information warfare and rumour conflagration represent a new set of hazards, the ability to defuse destructive waves becomes increasingly important. If Web 2.0 allows greater participation and influence by outsiders, responses by organisations to possible misinformation must also become effective, convincing and appropriate. Responses to outside feedback which are not honest, open or transparent will simply lead to greater mistrust and the erosion of a company brand.
Given the acceptance of blogs and the use of Internet blogs for corporate purposes, the move to blogs for internal information-sharing purposes seems a natural progression. Consider again issue development and escalation, but within the company, not just beyond it: there is a discernible unfolding of issues from latency, emergence, diffusion, maturity to reduction and closure. A firm might manage exceptions using excellent issue management and tracking systems, but firefighting is often a symptom of poorly designed processes, incompetence, bad product design and bad management. For a well organised, well run company, with a well functioning issue-management system, further business improvement can be gained by moving issue management to an earlier phase in the issue-management process. Blogs and forums can be used by experts and non-expert staff to raise issues (or monitor conversations for anomalies) while these are latent and isolated and have them brought to the attention of management. A specific example might be safety – noticing that a particular conveyor belt does not have a mesh guard – or trends with certain types of machine under certain circumstances – a machine from a particular manufacturer might have an exposed gearbox and be prone to wear and tear from dust. Blogs and forums provide a simple avenue for identifying and objectifying issues so that preventive action can be taken before accidents and breakdowns occur, in a sense providing a virtual ‘suggestion box’. And issue management is only one of many potential internal applications for blogs.
In cases where a firm produces goods or physical services, the company bloggers announce and discuss corporate wares and are often called ‘evangelists’, as the blog offers a kind of pulpit for their deliberations. Because these blogs can be linked to and commented upon, they exemplify the interactivity dimension of Web 2.0. Instead of organisations being closed shops and their designers, strategists and marketeers insulated from direct interaction, corporate blogs pull in ideas and comments from the entire congregation of interested parties. Although these corporate forums often provide a tightrope suspended over heresy, they are nevertheless aligned to and constrained by what the blogger chooses to articulate. Much of the language one reads in replies to blogs and discussions indeed suggests familiarity and virtual closeness: one part ‘the usual suspects’, one part ‘true believers’ and one part ‘concerned dissent’. Angela Merkel, the Chancellor of Germany, has an official video blog, but some students found it too unresponsive, so they created a website ‘direktzurkanzlerin.de’, where questions can be put by text or video and voted on by the public. After early resistance by the Chancellor’s Office, the demand became so strong that the top three questions are now answered each week (see Figure 2.4).
Figure 2.4 The student website which collects and rates questions for Angela Merkel Source: http://direktzurkanzlerin.de/
A similar function can be performed within an organisation. Formal and acknowledged experts can be given a blog, as can managers and administrators, from which events, concerns, product announcements or improvement ideas can be distributed and feedback received. While it is tempting to view this as just another broadcast channel for personal and managerial blogs in particular, the expectation in the time of Web 2.0 is that such communications be two-way, personal and authentic.
The next generation of blogging is so-called microblogging. This is the capability of blogging small amounts of text to a website in much faster cycles, typically from a mobile device. The site first credited with this innovation is Twitter.com, which allows text of up to 140 characters to be blogged. There are a number of functions available within Twitter which allow you to ‘follow’ a person by subscribing to their ‘tweet’ via a mobile phone, invite comments from people to your twit, and reply to tweets and so on. A Twitter.com address, like a standard blog, is generally personal, but there are interesting applications when the Twitter address is a theme, an event, a university course or some other ‘situational’ object. A lecturer could tweet their thoughts or delays in getting to a lecture: a mobile maintenance company could tweet news to its mobile fleet: a health programme in any large organisation could tweet reminders to its members to exercise that day. The Web 2.0 company Socialtext has a twitter-like product called ‘Socialtext Signals’ as part of its social software suite Socialtext 3.0 which is designed to provide this kind of contextual constraint to messages, allowing rapid, interactive communication while reducing the amount of signals and requests actually received by an individual.14 While interesting, however, this capability is already present in many standard applications which provide event-notification as part of standard workflow.
The HTML language standard was first conceived by Tim Berners-Lee, a British researcher employed at a European nuclear research centre. Working at the CERN laboratory, he was seeking a way for scientists to better share information and research results. He opened the world’s first website on 6 August 1991 at the address http://info.cern.ch (see Figure 2.5). As Attali says, many major innovations come from the work of publicly funded researchers who actually look into something utterly different to what they are commissioned to do.15
Figure 2.5 The first Internet website Source: http://info.cern.ch
A simple HTML page looks like Figure 2.6, where the words enclosed in diagonal brackets < > are called ‘tags’, which are interpreted by a web browser like Internet Explorer, Firefox or Google Chrome as commands to make a line appear as a heading or a particular colour.
But this language is not easy to learn and write. Tools have evolved to simplify the web page programming process and develop more sophisticated results. So instead of programming using this language directly into text editors, we use WYSIWYG (What You See Is What You Get) toolkits. Standard office products such as Microsoft Word or Powerpoint let you save a document as HTML, thereby allowing it to be opened and presented by the web browser. No HTML coding skills are required. Then there are sophisticated products like Dreamweaver and Websphere which are specifically intended for the development of powerful web applications. Although tools like this are a major productivity advance, three things are notable:
The process of publication is still document or file centric – the web page source is an object of some kind. It is developed and this developed instance has to be moved to a location where it becomes accessible to web server software
The process requires at least some level of skill which is not part of the portfolio of the normal knowledge worker. This includes page design at the very least and knowledge of stylistic standards, not to mention the mechanics of the products used to actually develop the pages.
The process of ‘promotion’ to a live web environment requires some form of administrative rights or privileges. It is usually not in the range of authorities for anyone to be able create and publish web pages at will.
In response to these barriers, wiki pages were developed in 1995 by Ward Cunningham, an American computer programmer, to allow the immediate creation and editing of web pages. A wiki user usually just clicks a function to request a new page which they can then edit and format. Typically, the user can create headings, tables or change fonts, upload videos and images and create links to other web pages or other wiki pages.
The wiki system produces the HTML code required by browsers, but generally stores this in a database rather than in individual text files. The HTML page is constructed and sent at the moment it is requested by a user at their browser from the data stored in the database. Changes to the page are stored in the database, along with previous versions of the page and the identity or IP address of the user who changed the page.
There are usually a range of standard functions in wikis for editing, linking, formatting, undoing changes, viewing the history of all changes, managing security and so on. In the final analysis, a wiki is a website (such as that in Figure 2.7) that can be dynamically constructed by users who have no knowledge of programming, but who do know, for example, how to do basic text processing. Early versions of wikis used a simple but nonetheless offputting wiki mark-up language to format wiki pages but most now provide a reasonably comfortable text editing system.
Figure 2.7 A simple wiki page using Mediawiki freeware Source: http://en.wikipedia.org/w/index.php?title=Transactive_memory&action=edit
An important standard function within wikis is the ability to define templates for different types of content page. When a new page is created, the template can be automatically inserted into the page with standard headings, boilerplate and even input fields with drop-down lists of allowed input values. This is extremely useful in corporate contexts: where consistency is important, direction for content creation should be provided, and although certain tasks might be routine, scope must still be allowed for variety. For example, a template for collaborative authoring of proposals and responses to tenders can automatically pull in company information, provide headings under which standard phrases are presented and so on. These templates provide important guidance to users and can themselves be adapted and enhanced by users.
Another notable wiki function is that of categorising (or tagging) pages: any page can be labelled with a tag which can represent an information category. When the category is searched for and found, all pages tagged with that category will be presented in a list. This resembles a keyword search with the added advantage that these categories can also be linked to each other in a meaningful way to reflect relationships in the real world. The category of hub cap might belong to the category of wheel which might itself belong to the category of chassis. But categories might also be less abstract: the category of Barack Obama might be useful to sort all wiki pages with some or any information about the president, but might also be linked to the category of Michelle Obama so that information about concrete instances of concepts can also be easily collected together and linked to information about other real-world entities. In Figure 2.8 we see how a conceptual hierarchy of category pages ‘Event’, ‘Machine’ and ‘Design Parameters’ can integrate information on a number of wiki pages, without those wiki pages ‘being aware’ of each other. A user will find these through navigating the category pages and then drilling down from a category page, such as Machine, to find relevant content pages, for example about the specific machine.
There are continual refinements to tagging in wikis: the Semantic Mediawiki group, for example, enhanced the base Mediawiki code to allow parts of wiki pages (as opposed to the whole page) to be marked up with metadata such that SQL-like queries can be run against the database. I might mark up a word in a page as a ‘German City’ and be able to ask the wiki how many German cities there are or to list all German cities mentioned in the wiki. A wiki page marked up like this:
There are dozens of wiki products available and most are available free to download or use directly as a service on the Internet. They vary in the power of their function, their scalability, whether they use a database, the integration of user enhancements and the brute number of users supported by the system. Some are more reliable than others, many are quirky (and proud of it) requiring users to be a little tech-savvy and prepared to rough it. To help the overwhelmed, there are several useful sources for comparing functionality and reliability.16 Some significant providers in the free wiki arena are Mediawiki, Wikia, PBWiki, Drupal, Twiki and Moin Moin. Most allow some level of HTML coding within a wiki page, but most also restrict this to prevent the possibility of malicious or dangerously incompetent code being introduced into the wiki.
If there are dozens of wiki products, there are literally thousands of public wikis on the Internet now. Special interest groups use them as a way to create collective knowledge and resources and generally have very low barriers to entry and contribution. There are several wiki indexes of public wikis, pointing to sites about law, philosophy, education or politics.17 Figure 2.9, for example, is a collective ‘Tax Almanac’ wiki which is not just itself an information source, but points to other wikis and information sites pertaining to taxation.
Figure 2.9 The Tax Almanac wiki Source: http://www.taxalmanac.org/index.php?Main_Page
The technology behind wikis is simple: the Mediawiki product, for example, consists of only about 35 data tables, although this has evolved over time to accommodate new functions (such as history, logging, content validation), enhanced performance and the growing volume of aging content. Unlike business process software, wiki technology supports general information sharing. It generally does not manage structured user data – the level of data granularity is at the page level, so there are no complex data relationships or minute adjustments continually required as business processes change.
One can divide the rest of the wiki products which cost money into two general groups: products which are available as stand-alone solutions and wiki functionality which is made available within broader-ranging products. These can be installed on a local server or hosted on a provider’s server and accessed as a service via the Internet.
The first set of products is exemplified by Confluence from Australian software firm Atlassian, and Socialtext. These are both highly rated by the Gartner Group and the information industry. In a nod to the open source origins of the concept, both make free versions available to individuals and offer a discount to community groups. Confluence resembles Mediawiki in its range of function, but is more robust, more user-friendly and corporate in its professionalism. It is also relatively cheap and has over 5,500 installed customers. Socialtext offers the usual functions of a wiki, but adds the ability to access it from mobile devices and to develop content offline and then upload this when a network connection is available again. It is also installed for corporations only as an appliance, that is as a physical server to be rack-inserted.
The second set of products covers those included in existing collaboration products or content management systems. Microsoft Office Sharepoint Services includes blogs, wikis, tagging and RSS feed capability as part of its overall collaboration and file-sharing suite. ECM, a leading supplier of content management systems, also includes blogs and wikis in its Centerstage Pro product. This uses the underlying versioning, security, information lifecycle management, archiving and retirement facilities of the Documentum product as a key selling point: the advantage of flexible, lightweight collaboration tools with the robustness, integration and control of a proven content management system. It is to be expected that any content management system will introduce the Web 2.0 capabilities into their function and that furthermore these will be ‘free’. Most companies prefer Web 2.0-type products to be integrated into existing content management systems.18
In the corporate context, wikis are used or intended to be used as knowledge and information repositories.19 They are flexible yet structured, persistent yet dynamic. They are particularly suitable for communities of practice who are developing knowledge or are sharing information pertinent to a discipline, but also have application as an intranet content management system, with the bonus of immediacy and interactivity. They are particularly useful for the distribution of answers to frequently asked questions and user manuals. Where blogs have a role in pushing emerging information, particularly to customers, wikis are most usefully the result of considered knowledge interactions and so not as suitable for customer exchanges, although they can be used for interaction with known suppliers and partners to exchange unstructured information in place of e-mail and telephone calls.
Corporate wikis will almost never be anonymous: any page created or changed will be associated with the user’s login identification. This of course will influence most people with regard to what information they are willing to enter into the wiki. Corporate wikis will also be user-constrained: the volume of people involved will generally be far less than if the entire Internet population could theoretically contribute: the ‘wisdom of crowds’ might be more like the wisdom of the one or two as firms generally keep specific expertise to the minimum necessary.
While case studies available on the intranet tend to provide success stories, these generally should be taken as demonstrating use-cases for potential application and not taken for granted as out-of-the-box applications. Organisations vary widely in their needs and in their ability to implement. McKinsey warn against the expectation that these systems will simply work. Indeed they record that about half of all adopters are dissatisfied and recommend a number of critical success factors, including leadership, support, integration into workflow, non-financial rewards and clear risk-balancing.20 Some early research analysis of corporate wiki use is also not exalted, although the evidence is limited. Blaschke found at an innovation agency which introduced a wiki to enhance collaboration that when a social network analysis was developed to account for indirect links and interaction on any specific page (rather than parallel wiki monologues), the amount of collaboration across authors was actually very limited and had to be judged a failure: he concluded that ‘the central concept of Web 2.0 was not to be found in the company’.21 Most interactions went through the main project manager who was also the wiki administrator. In seeking wiki-metrics, Ebersbach et al. found that in the first year of operation at Robert Bosch that a wiki, although producing many pages, showed little evidence of true integration or collaboration. For example, the low number of internal and external links and the different authors, versions and changes indicated that the pages were not being used with the fluidity and dynamism of which wikis are capable.22 Perhaps this is because in travelling lean, most organisations have as few duplicate specialists as possible so there is a dearth of conversation partners. Their conclusion, on the other hand, was that this revealed the ‘uncertainty and inexperience of the users’ (p. 153). Both these cases emphasise the need for further qualitative data to understand the reasons why wikis do not just ‘take off’.23
One of the miracles of the World Wide Web is not that we have so much useful information at our fingertips but that we find any of it. Amid the dross, the trivia and the nonsense, we still seem to find what we need. If you enter ‘IBM’ into Google, the first result delivered is the IBM home page. If you type in ‘Web 2.0′, the first result is a very useful introductory Wikipedia page. As you type in more keywords, so the result sequence homes in on your need. It is important to realise at this point that Google (or any other search engine) must maintain its integrity and deliver search results sequenced on an objective basis which has not been interfered with by Google (the famous ‘Do No Evil’ company motto). The first position or first page on a Google search is of enormous financial benefit and a whole industry of ‘search engine optimisation’ has been spawned by the need to make Google rank your pages higher than those of your competitors. Nonetheless, if users thought that Google could manipulate search results, for example to benefit a particular company, the search would lose credibility and usage would probably decline.
The miracle of search is due to the brilliant search engine logic which crawls all publicly accessible sites on the World Wide Web, indexing every meaningful token that differentiates one page from the others. Google’s patented page rank algorithm is of particular interest, as it is one of the key inputs to working out the relevance and anticipated likelihood of a page meeting your needs. As the web crawlers send back their information from all over the Internet, the indexing logic also records how many other sites link to this page. The more links coming into a page, the higher it becomes ranked in the search sequence. This is a manifestation of the ‘wisdom of crowds’ phenomenon – the more people consider a site to be worth linking to, the more likely it is to meet your needs too.
Finding what you need has become supported by increasingly sophisticated means, with search engines now building histories of personal and group searching in order to anticipate what result would mostly likely satisfy a request, building taxonomies on the fly which the search engine then applies to web pages to help you find them within ‘categories’ and so on. But finding pages is still challenging and if you have been researching a certain topic for more than a few minutes, even finding the pages you have already visited yourself can be difficult.
The phenomenon of social tagging emerged to address the difficulties of classifying and finding information. Put simply, it allows users to ‘tag’ or mark up a web page or document with a name which classifies that page in some way. These tags are formally called metadata, or data about data, which has been used to manage content in database management systems and information management circles for many years. They are social because they are available to others to use. The tags can be used to find information in future. You might create your own private tags or you might use the tags used by others. You might create a tag to describe web pages containing entertaining malapropisms or a tag called Beach Holidays for when planning your next getaway. These tags are created and kept in social tagging websites like digg.com or delicious.com. You simply go to one of these sites and enter your tag, which then becomes available for others to use too (hence it is social). You can search for the tag at the website, link it to other tags and add a web page address to that tag. When you look at the tag web page, it will list all the web pages and their links which belong to that category.
Physical objects described by physical labels are restricted by space, but digital signs are not and digital information can be described in many ways that do not interfere with each other. The classical methods of carving reality into ‘natural’ hierarchies (or taxonomies) which reflect an underlying essence are too restrictive, even if many categories do describe natural ‘joints’ in nature. Much of the world is in fact arbitrary and socially constructed and the capability of the Internet to allow people to tag, markup and share tags according to their various purposes is a useful advance in managing everyday information. In the same way as tagging, cooperative knowledge creation and information ratings lead to signage about the value of content. This is particularly important on the Internet, and to a degree in modern enterprises, where knowledge is a constant work-in-progress. Most significant news sites, for example, have a list of social tagging websites at the bottom of each page to help you note interesting articles for future reference. This has the useful side effect for the newspaper of making the articles more findable by others. In Figure 2.10, we see how The Australian newspaper, by offering readers the opportunity to tag articles, is simultaneously raising the connectivity of these articles and the likelihood they will be found by others. Die Bildzeitung, a German newspaper, offers social bookmarking and integration of their news pages with your personal blog (see Figure 2.11).
Figure 2.10 Referring readers to social tagging sites in The Australian newspaper Source: http://www.theaustralian.com.au
Figure 2.11 Referring readers to social tagging and blogging sites in Die Bildzeitung newspaper Source: http://www.bild.de
Here we see the ‘many hands make light work’ axiom in operation – but instead of many people creating information content (as is the case with Wikipedia for example), one is categorising web pages (working with information about information). And the more people do this, the more convergence and sense can be made of the vast quantities of Internet information. However, does this mean the categorisations are accurate or even correct? Some researchers take issue with social bookmarking in arguing that it not only degrades truth in classification, which is a normative act, but that it is philosophically unreliable: according to Aristotle, a thing can only belong to one category. If we take classification as an isolated, truth-functional process, this would seem to be correct: ‘Fido is a dog’ is true if and only if Fido is a dog. However, Fido is also brown, cute and the kind of dog I would consider buying: I could tag a web page about Fido with all these tags which serve my own purposes, and all can simultaneously be true. Further, classification is a final act on the way to making judgments: the construction of concepts is a social and permanently ongoing process. The consistent application of certain tags in clusters may lead to the development or recognition of new constructs. Tagging becomes a behavioural proxy for conversational objectification. Tagging overcomes the inherent ‘miscellaneousness’ of the Internet, allowing as many tags and sets of tags as there are points of view and purposes to which the Internet will be put.24
A particularly interesting component of tagging is ‘tag clouds’ which are increasingly a feature of search engines and large websites. These clouds consist of a display box of tags. Tags which have most referents or which have been most accessed recently (‘hot tags’) are displayed larger than less-used tags. Further, the tags which are displayed might be conceptually related to each other, so users can click on tags which they might not otherwise consider searching for. Figure 2.12 shows a tag cloud on the Blogscope website called ‘Hot Keywords’, where popular tags are displayed and magnified according to popularity.
Tagging has very interesting implications for organisations. It is an advance on document and content management systems which carry a ‘folder structure’ philosophy. According to this, a company department like Finance might have a document management structure which has a folder for correspondence in and one for correspondence out. But it would be useful to differentiate correspondence out to debtors, creditors, suppliers of raw materials and suppliers in a certain country: correspondence might belong to two of these categories but can only be in one folder. A single folder has as much meaning as the standard document management system can provide. But any number of tags can be applied to a document, as documents and web pages, by their very nature, often cover multiple topics and cannot easily be pushed into one category. These documents and web pages within companies become far easier to locate and retrieve when tags, and tag groupings, begin to emerge as part of business workflow. Personnel in different departments may know a document by a different name, or as a different type, or use it for a different purpose, or even use different sections. Tagging allows each group to mark the document using the appropriate vocabulary without affecting the location, structure or internals of the document or web page in any way. Furthermore, as user-created tags become highly used and institutionalised, they can be harvested and used as standard keywords throughout the organisation.
Syndication is a way of multiplying the impact of, and revenue gained from, information without a corresponding increase in effort. A syndicated news columnist, for example, writes an article once and it subsequently appears in many newspapers to which the columnist is syndicated. The same applies to television programmes, cartoon strips, lifestyle articles, horoscopes and recipes. RSS stands for ‘Really Simple Syndication’, a technology which uses the Extensible Markup Language (XML) to allow Internet syndication.
This system is a format which makes article headings, summaries and link information available on the Internet to any feed reader, a small and simple piece of software which runs in the background of Internet-attached personal computers. A person might be interested in current affairs and finance for example, and regularly browse the BBC and the Economist websites. These sites provide RSS feeds in special areas to publicise ‘what’s new’ to people who wish to subscribe. Users create topic areas in their personal RSS readers for news, sport or finance or whatever interests them and gather the web addresses of their favourite sites’ feed in these topic areas. The software then goes to websites at periodic intervals the user nominates and inspects whether any new articles have appeared. If they have, it displays a discreet message to the user and will download the overview, or header, information. If so configured, it can download the whole article or any embedded files such as video or audio. This latter capability is the basis of Apple iTunes for example, which will download podcasts from websites for music, language learning or radio series into the iTunes file system and then copy those to the iPod device.
Well-known free RSS readers which can be downloaded to run on your personal computer include RSS-Owl and FeedReader. A popular Internet-based reader is FeedBurner.com (now owned by Google), which gives you an account in which you set up your own feed reading preferences and publish feeds for others. Many related information management products include an RSS reader: Microsoft Outlook 2007 has a feed reader as a standard component, as do all commercial wiki products.
RSS is the glue that holds together the content systems such as blogs, wikis and web pages. It raises information to the level of salience, distributing targeted notifications to people and again overcoming the miscellany of the vast information resources of the Internet. At the enterprise level, some products (such as certain Mediawiki extensions for example) even allow subscriptions via RSS to a social tag or information category, such that any changes to any information in a certain class cause a notification to be sent to subscribers.
RSS helps to overcome the problem of information overload in organisations by allowing staff to receive notification of information about specific areas. So an engineer in a maintenance department might subscribe to the department manager’s blog for administrative announcements and to the local maintenance schedule page on the enterprise wiki. The engineer might also subscribe to the CEO’s announcement blog, the health and safety wiki page and the social club pages. The notifications contain links to information which is stored and maintained on a server. A job role, such as engineer or scientist or works supervisor, can be equipped with a default set of RSS subscriptions which tailors information notifications to those required by the role.
RSS is also of use at the project or ad hoc level. Notifications of changes to project pages in a wiki can be picked up by team members’ RSS subscriptions as can updates to discussion pages and special subject pages in which an engineer is interested or an expert. Every person sits in the centre of their personal information hub, and can tailor the information they are notified about to their own needs and according to their own work preferences. Notifications from an RSS reader can be inspected immediately or once a day, without disturbing a person’s concentration or work habits or clogging up their e-mail in-boxes.25
Since about 2005, there has been an explosion in the usage of computer websites which provide functions centred on the social self. Sites such as MySpace, Facebook, LinkedIn and Ning work from the inside out; that is, they provide facilities for people to describe and define themselves as a prelude to forming relationships, alliances and relationships using their self-constructed models of the self. Web 2.0 tools amplify the sociability of computer-mediated interaction through permitting interactions which resemble conversation: they are quick, they are easy, they are minimally structured, they are immediate. When a friend logs in it’s like they walk into the room. You can meet friends of a friend, pursue a chance meeting or just watch and listen.
Computer-mediated communities have existed since the beginning of the Internet in the late 1960s. Howard Rheingold describes the emergence of genuine groups brought together by shared interests and values in online conferences, bulletin boards and e-mail lists.26 The number of participants moved into the tens of thousands in the 1990s and although the Internet allowed the formation of relationships across the world, the patterns of communication and interaction were basically the same as those in the physical world, although the Internet is particularly strong in its support for the formation of ‘weak ties’.
The key tools of social networking are the social networking sites which act as hubs of activity. Self-disclosure is the first act, creating a profile of yourself which projects your personality, likes and dislikes, your interests and experiences into the common pool. You can then be found by others, usually your physical friends first, or you can invite others to become part of your network and sense their presence when they log in. Your network will grow or shrink, you will converse with some and not others and will discover new people who may or may not ever meet physically.
LinkedIn (with about 35 million members in 2009) is the pre-eminent hub for business and professional networking. Applying the same logic as social network sites, in LinkedIn, professionals can describe themselves, their professional background and qualifications, be recommended by colleagues or customers for further projects and join or form special-interest groups (see Figure 2.13). Interestingly, in the wake of the 2008 finance crisis, LinkedIn and Facebook both experienced greatly increased use as job-search mechanisms, which of course erodes their function as social sites and professional networks for knowledge exchange and partnership development.27
Figure 2.13 The Special Interest Group function of social networking site Ning Source: http://www.classroom20.com/?xg_source=ningcom
A highly-hyped platform for social interaction is Second Life, in which ‘avatars’, or graphic animations, can be used to represent an individual persona and conduct conversations in a kind of cartoon virtual reality. Use of this space is not restricted to entertainment: companies such as IBM and Ernst & Young, have set up shop fronts in this environment, and NASA have established a collaboration environment where scientists and team members can meet and converse. However, the use of this particular platform, in spite of the intriguing possibilities, already appears to be stagnating.28
There is a sense in which all tools which provide computer-mediated interaction can be characterised as social software, as the term ‘social’ applies wherever humans create and exchange meaningful messages based upon language and shared social structures. In a sense, asking Fred how many widgets were in stock was the face-to-face precursor of ‘social’ inventory management software (‘Stockbook’?). It is therefore useful to differentiate the Web 2.0 social tools from the conventional tools like e-mail or electronic forums, which have been used for thirty years or more. These tools can be characterised as ‘weak’ social networking tools, as they generally follow the patterns developed in non-computer-mediated interactions. E-mail, for example, will usually build upon existing patterns of communication and sustain roles which are developed in meetings, conferences or pubs. Similarly, electronic forums tend to be restricted to forum purpose (a professional project managers’ network of interest in German shepherd dogs for example) and so is constrained to these informational exchanges, albeit with humour and personal asides.
Strong social software allows the emergence of new patterns of behaviour and the definition, adoption and assertion of new roles that are not foreseen or constrained by the software itself. Strong social software provides an environment for the development of context over and above the exchange of information. The nature of this context (hostile, subservient, friendly or purposive) evolves through the use of the software.
Decisions about how, when, where and with whom to interact are individual and personal and express individual needs or desires. The less the decision is personal, the less social things become.29
The personal data entered become hostage to the intentions of the social networking site, intentions which usually involve the generation of profit. These intentions are constrained by a judgment of how far one can go with this data before the users object or feel exploited. Given that many people publicly document intimate details of their lives, one might assume that privacy, as previously understood, is unimportant. For example, in 2006 Facebook introduced an application called News Feeds, which automatically sent a message to a person’s friends when that person updated their personal data. Some 700,000 complaints later, it became clear that the problem was not about privacy, but about visibility – being in a mass of non-private individuals is fine until one is somehow lifted out of the mass. This may be analogous to the fact that while everyone knows that enormous amounts of information about oneself are stored, shared and collated by enterprises and governments, most people only care when it actually affects their next purchase or passport application. Or perhaps, as Montaigne wrote over 400 years ago:
Many things that I would not care to tell any individual man I tell to the public, and for the knowledge of my most secret thoughts, I refer my most loyal friends to a bookseller’s stall. (The Complete Essays III, Essay 9)
Similar applications introduced by Facebook in 2007 were Social Ad and Beacon. The former enabled Facebook to automatically send positive comments made by a person about a product to their friends and the latter revealed on the person’s public profile that they had bought a certain product. Both applications were implemented without consulting users – and the objections to these products were less about privacy and more about the unendorsed use of the information. This also seems to indicate a change in what privacy means, perhaps that the rights to a person’s information are now the main game – not the mere existence of the information somewhere in cyberspace.
In enterprises, there is also great scope for the application of social networking at the professional level and at the personal level. There are a number of software products which supply social networking functionality for use within corporations, and even one called WorkBook, which uses the functions of Facebook but stores the personal data on the enterprise’s servers within the firewall.30 Many organisations already use Facebook itself for their social networking, and there is much unsanctioned use by employees. This carries grave risks of breaches of security, disclosure of confidential information and the legal liability for publishing information purporting to represent the firm.
Large enterprises would probably benefit the most from internal social networking, as there are more possible configurations of groups and potential members who might not find one another other than through technologies which support group formation and connection building. Natural communities can form but, more importantly, personal connections can be maintained. Strong and sustained relationships are often formed, for example, between people in the same-year groups who are hired and inducted together but who subsequently disperse to other areas of the company. A company ‘Facebook’ allows these relationships to be maintained or at least rekindled when there is a need. People will often seek advice or help from people they have some affinity with rather than the official expert (who may be anonymous or even unpleasant), or ask their personal acquaintances for information on where to find an expert. A personal acquaintance will give information about a potential information source which is not necessarily publishable in a Yellow Pages directory of expertise.
An enterprise social networking site is not just a skills directory; it is for the enactment of discretionary personal relationships. Social software is a platform for enacting the sociability of the individual. If it is mandated, it just becomes a glorified business process or workflow management software. This is one of the institutional barriers that need to be overcome by new management approaches. Baby-boomer managers reject the very idea of pursuing personnel relationships using corporate technology at work, in the same way that baby-boomer parents cannot understand how their children can simultaneously chat on Instant Messenger, listen to their iPods and do their homework. Where wikis and blogs allow the publication of information or opinion to an open forum without management approval, social networking sites allow the formation of relationships independent of management fiat and of groupings which transcend organisational boundaries; this may threaten the power base of uncertain or vulnerable departmental managers.
The concept of the semantic web originated with the founder of the World Wide Web, Tim Berners-Lee. Recognising the rapidly growing volume of accessible information on the Internet, Berners-Lee proposed that a framework of machine-readable meanings had to be constructed to assist navigating and searching through the volume of documents, files and pages. He says the ‘… Semantic Web is … an extension of the current one, in which information is given well-defined meaning, better enabling computers and people to work in cooperation’.31 The intention is to allow searching, navigation and information retrieval to be performed by machines. HTML, the current dominant language in which web pages are written, is intended to allow a browser to present the information in a form which is easily readable by humans independent of the software platform: HTML markup tags are provided to highlight text, make headings and include images and so on. What are needed are mark-up tags which allow a software program to extract meaning from the information, not just the format.
The semantic web is an initiative of the World Wide Web Consortium (‘W3C’), the body which governs the standards and protocols of the World Wide Web. It specifies certain technologies for the task of creating and publishing ‘meaningful’ metadata (hence the expression ‘semantic’ web) which are designed to encapsulate all or part of the content of web pages, using tags which reflect meaning rather than format. These tags function as labels, as social tags or as bookmarks, but they allow a program to specifically interrogate a web page or document to see if it matches the desired criteria. XML (the ‘Extensible Markup Language’), while being a relatively mature technology, forms the basis of the tagging language.
RDF, the Resource Definition Language, is a modelling language based upon XML, which allows tags to be related to each other to form a data model which describes concepts, the relationship between concepts and the properties of concepts. OWL, the Web Ontology Language, is a more expressive modelling language than RDF and is proposed as the standardised ontology language of the semantic web by the Web Ontology Working Group of the W3C. OWL allows other features such as restricting the scope of the properties of a concept or the possible values of a property.
The essential outcome of the semantic web is rich, interconnected, standardised metadata which can be used to tag and add structure to documents, linking specific parts of their content to semantic maps or ‘ontologies’. An ontology is a conceptualisation of a particular domain of activity (such as dog breeding or conservation biology for example). The World Wide Web can itself be navigated by people or software programs until a target domain is found, at which point one ‘drills down’ to find concepts within the information objects in this domain. Shopping bots are an example of this: you may be searching for an ice-cream maker. A shopping bot site provides software search agents which scan the Web looking for this item. Within relevant web pages, standard XML data tags are found. For example, the price for the item, country of manufacture, guarantee period and delivery time are extracted and presented back to you, along with the result for other brands or stores, and you can then make a selection.
Alternatively, the objects, through association with the concept, are able to be filtered more effectively to provide a smaller subset of results from a search which more closely matches the searcher’s intent. For example, a semantic search could differentiate and filter results for ‘Jaguar’ based upon whether the search was for a cat or a car. This judgment can be made by a search engine based upon a user’s previous information behaviour. Having found the car, the user can find not only documents about the car, but also about motor cars (the class to which the Jaguar belongs) or wheels, chassis and engines, the parts of which a Jaguar consists.
The integrated knowledge map within the semantic web envisaged by Berners-Lee is an ‘ontology’, the network of related concepts which describes some domain of human reality. There are several ways to construct the semantic web or the analogues of it. One can deliberately create maps of concepts (such as taxonomies, glossaries or ontologies) which become a kind of controlled vocabulary and authoritative knowledge map which is used to classify information (for example by librarians). One can allow these maps to develop via organic growth and contribution, which is the approach underlying social tagging: groups of people unrelated in any way other than that they are interested in similar meanings can create or nominate ‘concepts’ and then begin to attach these tags to information. Concepts can also be created by software programs, which scan databases, documents and web pages and establish profiles of words, their context of usage and their relationships. Some search engines allow searches to be constructed and used by other users, effectively creating new shared concepts.
The notion of meaning is critical in the search function. An optimised search engine matches what the user ‘means’ with what the document ‘means’. Search engines now routinely construct user profiles from stored information (such as previous searches, browsing history, e-mail history and workgroup affiliations) and prioritise search results based upon the match of likely user intention with the meaning of the Internet or intranet information it has previously analysed and indexed. Indeed, through the use of agent technologies, users can be informed of the emergence of information which matches the users’ ‘meanings’. These agents monitor the Internet (or intranet) based upon previous searches which are either concepts or nascent concepts. For example, a search for uranium and Siberia in a sense is an instantiation of a new cluster of attributes which may be a formative concept (i.e. Siberian uranium): this ‘concept’ is then used (and even shared) by others to search or monitor changes in the information environment.
Meaning is measured by relevance within a domain of intention and action. So the fundamental intention of the semantic web is to lift information from the mass by its meaning and context. One application of this is that of ‘attention metadata’, where key elements are lifted from the mass to catch the eye of the user.
Within the corporation, metadata have been used for many years to describe database structures and documents. Increasingly, retailers use XML to make their databases available to shopping bots or even other companies to use in business-to-business e-commerce. But the semantic web moves metadata to an evolving, global platform. Specifically within the context of Web 2.0 ‘social software’ (as opposed to routine transaction processing for example), the notion of linked, navigable, searchable metadata relates to any content created in wikis or blogs or Twitter or the like. As we have seen, using some wiki products, such as Mediawiki, one is able to define concept wiki pages called ‘categories’. These categories are used to mark-up other wiki pages, allowing a search for a category to reveal all pages in that category. Further, these category pages can be linked to each other to form an ontology which defines an activity domain within the organisation.
So it is possible to build a map of the knowledge in an organisation using wiki categories and use this, at least, to mark-up wiki pages. Starting from a top-level concept like ‘Our Business’, one decomposes into the top-level business processes, sub-processes and then concepts which are relevant within the sub-processes. This tree can be displayed using standard wiki functions or placed within standard wiki information pages using the Mediawiki CategoryTree function.32 This is an example of a local ‘semantic web’.
Tagging is of course not restricted to documents. On discovering that 40 per cent of employees had not updated their personal pages in the previous nine months, IBM developed a social system called ‘Fringe’, which was based upon the data from the standard personal pages with some enhancements: in particular, it allows IBMers to tag their colleagues in much the same way as social bookmarking tags websites. This increases the specificity and currency of metadata about a person’s capabilities by using the ‘wisdom’ (and energy) of crowds.33
Proponents of the semantic web themselves remain agnostic about the prospects for its success. A semantic web for an already rigorous and normative area such as medical science or geophysics is strongly facilitated by the momentum of the discipline. But as Antoniou and van Harmelen say ‘… the greatest challenge is not scientific but rather one of technology adoption.’34 Who will write the ontologies, who will authorise one ontology over another and where will the semantic mark-up of documents come from? This is particularly pertinent for ontologies to be used by business-to-business transactions, web services and search engines. These authors surmise that ‘the first success stories will not emerge in the open heterogeneous environment of the WWW but rather in Intranets in large organizations … we believe that knowledge management for large organisations may be the most fruitful area to start.’35
But the business case for such detailed mark-up is ambivalent. The precursor technology to semantic mark-up called SGML (Standard Generalized Markup Language), which allowed organisations to create their own tags, has enjoyed moderate but not resounding penetration. To generate a return on investment for semantic mark-up that goes beyond high-level tagging or categorisation of documents, an organisation would have to generate a very large number of (probably) complex yet structurally consistent documents. The general mode of processing those documents would have to rely upon a high level of specificity (for example, find me all financial documents where the County = Devon and the revenue from dog licences was greater than 500 pounds). Hansard (the proceedings of Parliament) is a good candidate for semantic mark-up, as are insurance claims or the documents created by journal publishers.36 But until the cost and effort come down (which of course they will), most organisations would likely be satisfied with a higher level of granularity, that is being able to navigate through an ontology and find documents which fall within a category.
A mashup is a composite web page, one that consists of parts drawn from information facilities on another web page or provided by services on a website. Individual websites usually have a competency in a certain type of content or a mode of presentation. Real estate agents know where houses and apartments are for sale, government land planners know where new highways will be built, the police know the distribution of crime and house burglaries and a bureau of meteorology has the weather statistics of the past 100 years. These are examples of where the core competency of a group becomes explicit in the form of structured data. Google Maps is an example of a competency in representational form, namely two-dimensional mapping as well as three-dimensional image rendering of streetscapes. An example of a mashup is to bring housing, crime and planning information from different websites into a geographical layout using Google’s mapping facilities, such that people can make informed decisions about where to buy a house using specific data to answer the questions people usually ask.
Such a product might be for a particular niche of users and not be worth the effort of data collection, organisation and programming. However, if the effort is limited to building dynamic interfaces to those websites and then combining that information with standard Google maps, then one has a cheap-to-build service that may generate some revenue or sufficient social benefit. A neat example is the ‘unfluence’ website, which is a mashup that maps political donations to congressmen (see Figure 2.14). (For a similar idea, see muckety.com, which is a website that publishes schemas of relationships between individuals and institutions along with its news stories.)
Figure 2.14 The Unflence mashup of political influence on US politics Source: http://unfluence.primate.net/
The key to mashups’ capability, particularly in enterprises, is the construction and publication of web services via a services-oriented architecture (SOA). An SOA is a blueprint for how an enterprise will deliver data using new, powerful techniques of programming, design, networking and networking standards.
A cursory examination of the limits of conventional programming structures is apposite. Using standard relational database programming methods and languages such as the Structured Query Language (SQL), data can be retrieved from databases and presented to both users and programs for display or processing. One can easily combine data, filter it and perform basic calculations using SQL. However, this is a low level of ‘value add’ to the data and corporations need more complex processing involving validity checking, combination of data across different systems, sophisticated calculations and so on. This is currently solved by application programs for accounting, finance, works management, logistics and so on, which deliver ‘screens’ of information and which are driven by user functions to add an order, enquire on the levels of widgets in warehouses across the organisation or in a certain area, or compare the budget versus actuals for fuel and the transport of goods. These applications are delivered overwhelmingly via commercial off-the-shelf packages and often integrated as modules into enterprise resource planning suites. The best known of these is SAP’s R3. While these are powerful systems that are flexible and configurable, there are some disadvantages:
Not all required data is presented on one screen, requiring navigation and swapping between applications, for example if one wanted to see the number of widgets in the Spandau warehouse, their financial value and the value of those on order.
To redesign screens, even slightly, to fit in data which is required by a new business process requires reprogramming (which only the package supplier or the information technology department can do). For example, a sales screen might need a field which states the creditworthiness of the customer, but this simple Yes/No might be the result of a complex process of combining previous history with Dun & Bradstreet credit data.
Any project which is seeking to develop a system for a different business area and which uses portions of the existing data must develop its own programs to access and process the data for presentation. For example, a product support system for a new call-centre will need to access existing customer and product data, as well as warehouse data.
The underlying conundrum here is perennial in information systems: reuse of data and programs, and how to design, package and deliver software which, like Lego bricks, can be combined to build whatever structure the builder requires: write once, use anywhere, use often.
SOA is an approach to building reusable program components which deliver data and make them part of an organisation’s infrastructure.37 The components are discrete, visible, documented and have a clear, single purpose. They must be written in such a way as to be easily orchestrated into an ensemble which fulfils a business need. And the data services provide a single point of truth – there are no competing versions or combinations of data where different programmers have performed slightly different operations on data to attempt to achieve the same goal. On the Internet, SOA services are better known as widgets and a good place to observe these is in iGoogle, where a user can decide to place SOA components which deliver the joke of the day, the time of day or a favourite news feed.
From a set of SOA artefacts one should, with little effort and possibly no programming skills, be able to construct a screen which supports a particular business act. This capability can accelerate the construction and delivery of new information systems and applications but also fill a niche for what IBM calls ‘situational applications’. These are applications which may have a shorter life span than usual (a few weeks or months even), for which scarce and expensive IT programming resources are unavailable and which perhaps do not generate a sufficient business case.
Currently, many large organisations have recognised the potential productivity gains from this kind of programming ‘holy grail’. Now that this is technologically feasible, SOA generally exists as a plan, or program of works, which revolves around enterprise architecture and the disciplined design, deployment and exploitation of the data services. Many large packaged software providers (such as SAP) have promised to deliver their functionality as SOA components so that customers and users of their packages can use them directly for new systems or situational applications.
This sounds of course like software engineering, not Web 2.0: why is this mentioned at all in the discourse about Web 2.0? The reason is that Web 2.0 exemplifies the consumerisation of SOA applications: mashups based upon the ability to easily combine functionality into a single screen are an example of how modern Internet users interact with and drive the systems themselves rather than passively accepting whatever is available or served up by the IT department. SOA will be a driver of business information productivity, whether or not Web 2.0 exists, but making it available and easy to configure at the user interface will be demanded by users – and will indeed accelerate productivity increases based upon information use.
A good illustration of SOA meeting Web 2.0 is QEDWiki, or Quick and Easy Development Wiki, first developed as a product prototype by IBM and now existing as LOTUS Mashups. This integrates wiki technology (immediate page writing, discussions, tagging and so on) with the ability to compose business screens by combining SOA component services. This is a non-technical process, supported by a drag-and-drop interface, for ‘power users’ or business analysts. Significantly, this screen is then available to any other users of the wiki, so a team can create and use for a brief or extended period the wiki part for information interactions and the SOA parts for data provision and processing. This screen can be used by others in the future, creating a powerful network effect as word about the screen functions spreads. Security is not the concern of the user, because this is managed within the SOA component as it accesses relevant databases.
Sandy Carter, an IBM Vice President of SOA, predicates the implementation and success of Web 2.0 in enterprises on flexible yet standardised data services provided by SOA, and certainly routine business processes where structured, accurate data are provided are where the rubber hits the road. She cites a survey in which 80 per cent of CEOs identify rapid response to changing conditions as a major competitive priority, but only 13 per cent rated their organisations as ‘very responsive’ – indeed the majority of CEOs perceive their IT departments to be inhibitors of flexibility.
This is indicative of perhaps two things. Firstly, those business transactions are at the heart of commercial activity and that data is a primary driver of business transactions. This is what needs to be managed first and foremost in order to deliver products and services according to a controlled regime, but it needs to be managed flexibly. A McKinsey 2008 Web 2.0 survey showed that the most important ‘Web 2.0’ technology for business users is web services, the use of SOA to deliver packaged business function. Fifty-eight per cent responded affirmatively to this technology (oddly down from 70 per cent in 2007) compared with the next most used technology, blogs, at 34 per cent. This suggests strongly that delivering specific, data-driven business functionality is the focus of organisations and has a higher priority than free-form information exchange made possible by other Web 2.0 technologies.
Secondly, it is perhaps indicative of the interests of companies like IBM who are providers of programming services and technology. Web 2.0 products like wikis and blogs are not going to generate much wealth, either in software or consulting fees.
Web 2.0 implementations in the enterprise usually involve individual products or several products together to achieve some sort of coherent and synchronised informational flow. Indeed, it is no accident that most products offering wiki capability also contain blogs, RSS feeds and tagging as a minimum. It is quite simply because these products complement each other in knowledge management:
The use of one leads naturally to the other – an entry in a blog or wiki should trigger an interested party to go and have a look (RSS). An interesting post in a blog should be recommended to others and an article about bird manure fertiliser should be tagged ‘guano’ even if that word is not in the article.
An environment in which one of these tools is used implies fertile ground for the application of others. Consider the extensive use of Web 2.0 tools by the USA’s top 100 newspapers in Figure 2.15. These facilities are combined to provide readers with a complete functional environment for interacting with the newspapers’ information.
Figure 2.15 Use of Web 2.0 facilities by US newspapers Source: http://www.bivingsreport.cam/resources/2008.gif
A convergence of many forms of software is taking place; as the overlaps between document management, forums, wikis, blogs, RSS and e-mail become clearer, so vendors will offer all these functions in a single suite. But why do these products converge? It is because knowledge processes cannot be constrained to a single form of interaction. Any one of these tools on its own does not suffice – taken on their own, they will lead to fragmented and undisclosed stores of knowledge. That is why the products must be viewed as supporting an overall system of knowledge transformation. Further, these products operate at two levels of meaning. Where wikis, blogs, mashups and social software contain or present content, social tagging, recommendations, ratings, the semantic web and meaning-based search operate on metadata, making the content salient and visible.
When implementing or designing structured business applications, one usually develops business process models and flow charts which represent business activities and tasks. Then one identifies what pieces of information or knowledge are required to accomplish the specific task. But with collaboration and exchange of unstructured information, the paths are not always formal and predetermined. Sometimes exchange and collection of information can be steered using so-called workflow products which will route documents or text from one designated person to another (for example, via e-mail), capturing their changes, reviews and approvals. But when it comes to collaboration and innovation, big ideas can come in small packages and progress is often dynamic, nonlinear, recursive and complex. Small interventions can create huge opportunities (a single brilliant word) or destroy them (the manager entering the room). Value in collaboration and knowledge sharing is not related to quantity, duration or routine.
Figure 2.16 shows how the Web 2.0 technologies can be combined to deliver a suite of open-ended services to an individual information worker. The worker will adapt these to suit their own purposes: the services are common, general-purpose infrastructure and can carry any type of information to any required level of detail or granularity. These services include:
Figure 0090.2 shows how individual knowledge workers, through the intermediation of various Web 2.0 components, are linked to each other and have facilities to capture, describe, find and share knowledge through wikis, blogs, social and ‘semantic web’ tags, RSS, social networking services and mashups.
3.Enterprise 2.0 is the term coined by Andrew McAffee of Harvard University (McAffee, 2006). It has gained prominence as a term building on the concept of Web 2.0 to describe enterprise use of Web 2.0 technologies according to a certain profile. While this book is about this theme in general, I have preferred to steer clear of using the name, as it has spawned its own controversy and taken on a trajectory of its own. For an excellent review, see Buhse and Stamer (2008), which describes the use of Web 2.0 tools in organisations.
4.‘The Art of With’, a seminar held in Manchester on 24 June 2009.
5.The Technorati website tracks blog use and activity and their blog survey is available at: http://www.technorati.com/blogging/state-of-the-blogosphere/.
7.See ‘Robert Scoble on Corporate Blogging’ at: http://www.cioinsight.com/c/a/Foreward/Robert-Scoble-on-Corporate-Blogging/.
8.For example, according to the Bivings group, 93 per cent of the top 100 US papers have reporter blogs (http://www.bivingsreport.com/2008/the-use-of-the-internet-by-americas-largest-newspapers-2008-edition/).
13.For the latest numbers and trends see the Technorati website: http://technorati.com/blogging/state-of-the-blogosphere/.
18.Anonymous (2009a). Further, integration into current suites of content management software increases manageability of unstructured and semi-structured information and documents and increases levels of control and oversight.
25.Some studies show that e-mail arrivals, even when corporate spam, cause a significant loss in concentration and work momentum.
27.‘It’s a safe bet that if the economic downturn grinds on, we will witness further conflict between the nonrational instinct to connect socially and the rational calculation to build social capital for professional reasons. If so, it may put further strain on the notion of an online friend. We may find ourselves asking more frequently that age-old question, “What are friends for?”‘ (Dutta and Fraser, 2009).
32.This is described in the section on wikis in this chapter.
36.For example, the Canadian government makes up its proceedings of parliament using semantic mark-up, as do many other countries now.
37.The definition of SOA from IBM is as follows: ‘A business-driven IT architectural approach that supports integrating your business as linked, repeatable tasks or services. SOA helps today’s businesses innovate by ensuring that IT systems can adapt quickly, easily, and economically to support rapidly changing business needs. It is a flexible architectural style that enables customers to build a set of loosely coupled services for automating and streamlining business processes’ (Carter, 2007: 288).