I had a great lunch with a friend of mine, Martin Garland, yesterday to get caught up on what he's been up to. Martin is really interesting and has his hand in several areas of the ECM business including heading up conceptSearching which provides some very interesting search technology Some of the capabilities provided include the following:
- Probabilistic Latent Semantic Indexing
- Relevance Ranking based on Probabilistic Model (Bayesian Inference)
- Concept identification based on Shannon's Information Theory
- Cross platform compatibility via Web Services All Application Programming Interfaces (APIs) based on XML
- Transparent access to system internals including the statistical profile of terms
This technology becomes very relevant when you start looking at the exponentially growing quantities of unstructured content and the difficulty users have when attempting to find information. Traditional structure style Boolean search expressions begin to break down when looking for patterns within documents. The difficulties are compounded if there is uncertainty about the presence of documents and the exercise is designed to gather evidence, or to prove the absence of, information about the selected topic.Things can be improved when adding metadata to provide context for the documents and how they are used within the business. One of the new products that I plan to get a demonstration of Monday that really excited me is their conceptRuleBuilder. conceptRuleBuilder offers users a web based interface that enables domain experts to easily review and maintain taxonomy nodes, without the need for IT literacy. By establishing a taxonomy that reflects the way the organization is using information search accuracy is improved and the way users interact with the information can be improved.
I will provide an update after I get an updated demonstration of both conceptSearching and conceptRuleBuilder.