Knowledge Management and Portal Technologies

In parallel with the evolution of portals in law firms was the evolution of search. At first primarily a facet of the document management system, this DMS search was poorly used for three broad reasons. First, the scarcity of the solutions themselves ' many firms disabled the functionality for that reason alone. Second, because of the 'signal-to-noise' ratio: unless the firm had the foresight to impose and enforce classification schemes, searches returned few useable results sets. Last, the complexity and number of search syntaxes meant that queries were often too broad. Also, in most cases, no context was captured during the process of saving to the DMS, so the documents returned were, at best, information not knowledge.

These engines were keyword based, with some offering basic word stemming and thesauri. More advanced search solutions began to evolve that offered auto-classification and entity extraction (e.g., the identification of which jurisdiction a contract is written in). These early frontrunners leveraged pre-canned queries built by experts run using a keyword search engine (e.g., Microsoft SQL search). These solutions also offered search over a number of repositories ' the DMS and e-mail, for example. A similar ASP-based service was provided by LexisNexis. While this approach continues to be effective, it tends to require 'tweaking' of the rules used to analyze and classify it (if it is to be effective in more than 70%-80% of documents). In addition, it is not easy to change the taxonomy(ies) supplied as new (lengthy) queries need to be written. Nor are they necessarily suitable for multi-lingual practices unless complex translation/synonym lists are used or, more accurately, the rules rewritten to account for changes in language, grammar and parlance.