So, what's a "Capability Case" and why should I care?One of the things that we're doing with Project Shelley (and all of the OpenQabal projects, really) is expressing the initial requirements in terms of capability cases. That link explains capability cases in more detail, but the gist of it is this: A capability case is a business problem, linked to a set of technological capabilities, through a scenario. A capability case could be considered somewhat similar to a use case, but capability cases are more specifically about linking the scenario to a business problem and envisioning a solution, expressed as required capabilities. To illustrate the point, and to get the ball rolling with describing Project Shelley in terms of capability cases, here's our first Project Shelley Capability Case. (pdf), (odt).
Collaborative Filtering for Information Retrieval
Use voting/ranking by individual users to tap into the “wisdom of crowds” effect to filter / select the most relevant information in a given context.
Knowledge workers - and especially executives - face nonstop demands on their time, and have to make key decisions in ever decreasing time spans, in order to adapt to the rapidly changing business environment. Balanced against the need for rapid decision making is the need to consider and evaluate as much available information as possible before making a decision. The information needed to make correct decisions often exists, either within your enterprise - often locked away in the collective, accumulated wisdom of every member of the organization - or somewhere outside your enterprise. In either case, it can be nearly impossible to solicit the correct information before risking a strategic mistake. In any medium to large organization, it simply is not possible to review every document and poll every employee, customer, partner and vendor before executing a decision. Even if time were available to do this, a small nugget of essential information could easily be lost in the sea of noise. However, technological tools make it possible to rapidly filter, rank and correlate various sources of knowledge, helping to ensure that what is important makes it to the people who need it, despite it’s origin.
A group of individuals can often be “smarter” than any one member of the group. By aggregating the wisdom of individuals via voting / ranking / correlation using collective intelligence it is possible to tap into the wisdom of crowds effect within your organization. Collective intelligence ensure that relevant information is seen by those who need to see it, even if it “bubbles up” from an otherwise obscure source.
At MegaCorp, the worldwide leader in enterprise software with their flagship Flozzit product; sales were down and managers were scrambling to increase revenues. A group of managers decided that the solution was to create a new, feature-enhanced Flozzit 3.0, which would add missing features and solve long-standing issues that were resulting in lost sales.
Lacking sufficient resources to add every desirable feature, it was critical that the Flozzit Product Manager identify the features which would most directly impact sales. So the Product Manager began scouring over enhancement requests in the bug database, and scanning old emails from account managers, field reps and engineers. After a few weeks work and several meetings, the PM thought she had a pretty good handle on which features should go into Flozzit 3.0.
Before committing resources to the new roadmap however, she decides to peruse the “Flozzit” channel on the PS portal, and look for items tagged “complaint” or “enhancement.” At the top of the list is a report written by a customer support representative (who the PM had never met, or even heard of; she wasn’t even sure if he was still with the company) titled “Why BigCorp hates Flozzit.” Intrigued, she examines the filtering metadata and sees that nearly every CSR in the company has upvoted the report, as well as one or two of the engineers. She downloads the report and digs in, to find a detailed summary of the top issues that end-users at BigCorp (the largest customer of Flozzit!) had complained about when talking to the CSRs. The language was detailed and some of it was not kind to Flozzit. After reading the report, the PM arranges to meet with the CSR who wrote the report, and identifies 5 top issues which had never been discussed in the many meetings held to identify the new Flozzit 3.0 roadmap. She then calls her top contact at BigCorp to discuss the issues and the first thing he says about issue #1 is “Yes, our users have been very concerned about that. We noticed that HyperCorp is releasing that feature in their 4.0 product and might consider switching if MegaCorp doesn’t answer soon.”
Armed with this new information, the PM polls a sample of other Flozzit customers about the 5 issues identified and find that 3 of them are so important that they must go into Flozzit 3.0.
Six months later Flozzit 3.0 ships with the 3 new features and a slew of bug fixes. BigCorp immediately commits to an upgrade, and are so happy with the new version that they purchase another 50 licenses a few months later.
Corporate culture which stifles dissent.
Lack of incentives for participation.
Lack of belief in the utility of the system.
Unsupported document formats, databases with proprietary formats which are difficult to
Better identification of actionable news and information which might otherwise remain
lost in the sea of information inside the enterprise; leading to better decision making at
both the strategic and tactical levels.
Share links to web-sites, documents and other items of interest
Categorize links by topics using channels
Tag links with specific keyword
Rank items by voting them “up” or “down”
Search and filter by topic, keyword, and/or score
Sort view by various statistical measures, such as “all-time score”, “hotness”, and
Typical Use Scenarios and Guidance:
Knowledge workers view channels of topical concern to their jobs, or of general interest, on a regular basis, voting and tagging existing items, commenting on existing items, and submitting new items, creating a view of what’s important - and adding to the corporate memory - using collective intelligence.
Knowledge workers discover relevant information through casual browsing; and through directed searching by tag, channel, submitter, score, or other attribute, when specific topics are under review. By limiting causal browsing to the most highly ranked items, an employee can maintain a “finger on the pulse” of what is considered important at a point in time, wthout reviewing every item. But directed search makes all of the other items accessible when they are relevant to a topical query.
Fogbeam Labs “Project Shelley”
Other corporate knowledge repositories (blog servers, forums software, document
management systems, HR management systems, etc.)
Existing Data Warehouses / Databases / Knowledgebases
External information sources (web pages, databases, etc.)
Project Shelley can easily integrate any knowledge source which can be accessed via
HTTP and which exists in a format which can be parsed into text tokens for indexing by
Lucene. Where text extraction is not possible, location through metadata is still possible (ex, mp3 audio files, video, etc).
RSS feeds, HTTP, OpenSearch