One of the key tools in any truly knowledge-managed organization is a business analytics feature set for mining and analysing text. Attensity recently updated their line to version 4.1, and the results are impressive. Text mining tools plumb unstructured and structured documents to find and expose relationships between various kinds of data. If the tool supports mapping of concepts and vocabulary, it can focus d Text mining tools plumb unstructured and structured documents to find and expose relationships between various kinds of data. If the tool supports mapping of concepts and vocabulary, it can focus discovery in areas like product failure analysis, fact-based marketing clustering, and competitive intelligence as well as fraud detection. The more text an enterprise is larded with, the fewer clear possibilities most human analysts can latch on to and the more relationships a text analysis tool can condense and offer up for the analyst’s consideration. Attensity calls their multi-module text analytics product a “suite,” and while I’m not ready to assign a specialized tool such a vaunted noun, the five browser-based functions do complement each other pragmatically. “Text Search” supports iterative exploration of text documents. “Discover” supports categorization, combining, and clarifying words, as well as structuring results for further exploration. What Attensity calls “Analytics” is a module that delivers graphical visualization tools to examine results (I have to say I really dislike the trend in BI where companies call graphics “analytics”). The “Alert” module triggers messages based on emerging conditions in data, allowing rapid event response. The interface is fairly typical for text analysis tools: neither elegant nor confusing. An analyst working with the various modules will find a real but shallow learning curve to cope with. Attensity 4.1 targets both data extraction (which operates in response to the organization’s prepared knowledge engineering specs) as well as “exhaustive extraction” which seeks relationships without any specs, a significant potential benefit. The interface is fairly typical for text analysis tools: neither elegant nor confusing. An analyst working with the various modules will find a real but shallow learning curve to cope with. Attensity 4.1 targets both data extraction (which operates in response to the organization’s prepared knowledge engineering specs) as well as “exhaustive extraction” which seeks relationships without any specs, a significant potential benefit.There are a few features not yet in Attensity’s system that I’d like to see added. For one, I’d like the system to learn to distinguish ambiguous “hits” that fail by having the user identify them as “misses.” And while Attensity can direct output for reporting in Business Objects’ Crystal Xcelsius, it doesn’t currently connect effortlessly with more compound Business Analytics solutions from SAS and SPSS (which have their own text mining tools).Text mining is one of the vastly under-appreciated 21st century applications. Organizations that hop on this set of capabilities will find advantages over their competitors that are not soon evened out. Attensity 4.1 Platforms: Runs on Windows and Linux; works with SQL Server, mySQL, Oracle, or Teradata source databases Cost: Solutions start at $250,000 for the server application license; a hosted application suite starts at $15,000 per month Verdict: Attensity 4.1 is a cleverly-formulated competitor in a vital but generally ignored category: text mining for knowledge management and business analytics. Its capabilities cover both targeted extraction (where you know what you’re looking for) as well as untargeted (where Attensity discovers potential insights you may not know about). More prepared connections to popular platforms that would benefit from Attensity’s features should be built. Technology Industry