Enterprise Search and New Idea Engineering (and see taxonomy methodology). Raritan's Topic Explorer. The IRS tax map (lots of screen shots, algorithm, light on method). IRS tax map (2007; strong on method and project issues); IRS tax map (2003; strong on early issues with an indexed approach). Usability testing for IRS tax map. Testing taxonomies for usability. Card sort as a taxonomy development method; Boxes and arrows on card sorting. Adding network analysis to card sort.
The old call center use case for KM.
Important insight: Store failed searches, they are often proxies for subjects that are stored under different proxies. Then feed the relation between a successful search and the failed searches back into the system.
The long tail and the short search box. 2006, but still seems true. (SOLR faceted search best replacement for "advanced search" that I have seen, and being adopted everywhere.) Domain vs. technical expertise quadrant for search. "Novices Orienteer, Experts Teleport." Faceted breadcrumbs -- really nifty! (And after "advanced search" there's "guru search" -- enter triples in the box!.
Jakob Neilson (2009) on megamenus, which all the kool kidz are using.
More from Neilson: Is search so good it makes us stupid? (How can we encourage users to change search strategy for find the most useful results as opposed to the first results? One answer is to know more about the user, hence cookies, and so forth.)
A semantic text search engine for categorized data, such as varchar fields from a database (probably like most – if not all – data in your application).
Forester on enterprise search trends (2011). WARNING: Microsoft sponsorship! The abstract:
First-generation enterprise search was not easy to use and produced unsatisfactory results from a usability and relevance standpoint. Today's knowledge workers demand role-specific, contextual search everywhere they work. In recent interviews with Forrester, search experts expressed optimism about their ability to deliver search solutions that are both usable and useful. Search technology is mature and stable; for the most part, bottlenecks regarding scale, security, and connectivity can be easily resolved. So the trends to watch focus on strategy, not technology. In 2011, content and collaboration (C&C) professionals responsible for search will position themselves as trusted advisors to the business, add value to the social output of empowered employees, and confront content quality problems head-on.
Content quality problems, eh?
Visual search. Meh. Custom.