The New Frontier—Capturing User Context
The platform imperatives include adding contextual metadata into the search query about the user’s current and past work and current application usage, in addition to their role, interest and interaction history. Attaching such fragments to a search query gives the back-end that much more data to work with. Not only do you have all your old metrics about the indexed content, you can also pull out all the metrics available on the user context to predict accurately the user’s intentions.
By putting all this information into a topic map-like structure, and building a similar topic map on top of your indexed data, you enable the finding of similarities and links between the front-end and back-end topics to boost relevance. You will end up with links between documents, authors, roles, departments, products, dates and times, one-way links, two-way links weighted differently each way, access rights, viewing logs. etc. The search terms will be your primary picker for valid candidates, and the topic contexts for the document and the contexts of the query will give you the valuable information on how to rate this particular information.
Next is the presentation of dynamic results. This is the brokering function of the search platform. Search platforms will provide result-list interfaces with dynamic mash-up presentation of data elements and application components dependent on the user context—all being configurable by the customer in relation to the end user’s role and interest. We can already see the adoption of this technology in large consumer portals, and expect this to become prevalent in enterprise search also.
With this technology, we expect the emergence of innovative and search-driven end-user applications. One area is the usage of automatic searches that use only the user context without an end-user search-term entry. The application IntelliSearch Knowledge Assistant is an example of this. While writing this article, one of the authors used this application and switched between Microsoft Word and Outlook, reading emails from a colleague about this article. Doing so, he was alerted by the Knowledge Assistant of a relevant document on the topic of contextual search, located on our intranet.
Customer Imperatives
The opportunities that "search 2.0" represents have many implications for customers. For one, they will have to engage with the vendor in determining contextual design parameters and mash-up techniques when requesting functional specific search solutions. Secondly, they will have to standardize on an application-independent search solution in order to capture the user context across the desktop applications. This entails inducing users away from the application-specific search box, toward a central search solution located either through right-clicking on the mouse, or as a desktop search query box in a taskbar. The search box belongs to the user and the desktop, not the indexed repositories. Finally, customers must be willing to experiment with the technology in order to unlock its full potential.
What’s New? IntelliSearch Knowledge Assistant
The Knowledge Assistant continuously monitors, finds and alerts information workers of topical and textual documents similar to their current work in Microsoft Word. The Knowledge Assistant is both an ad-hoc enterprise search tool and a monitoring agent. You can use it to search for files or to monitor areas of interest from any source, internal or external. The Knowledge Assistant will automatically search and find data relevant to your current work. E.g. while you are writing a document in Word, it will automatically search and find other documents, files, internet articles, etc. containing relevant information according to what you are writing.
Your Personal Knowledge Assistant—search for relevant information while you work
The Knowledge Assistant comes integrated into the system tray/task bar. The solution supports Microsoft XP and Vista Office versions. The Knowledge Assistant is part of the IntelliSearch Enterprise Search platform that received the "KMWorld Trend-Setting Products 2007" award.