Automating Perception Article from KMWorld Magazine


Abstract
An article from the April issue of KMWorld magazine looks at three companies in the information retrieval space (also referred to as categorization and classification). One of those companies is well known to patinformatics practitioners, ClearForest, while the other two are less well know names.



For a nice brief article on new categorization and classification technology have a look at the Automating Perceptions article in the April issue of KMWorld magazine. A link to the full-text can be accessed here . The author prefers the term automating perception to any of the other more common terms so he uses it throughout the article. Personally I think the term is a bit soft and doesn't really convey what these particular types of systems do. I much prefer terms like information retrieval or information extraction. The three companies mentioned are Convera, ClearForest and Stratify. Convera deals with document categorization and taxonomy building while ClearForest is more concerned with inner document retrieval of data nuggets. These days the company tends to call this process SmartTagging based on the current interest in XML but I have always preferred the term information extraction since the software examines unstructured text for key pieces of information and puts them into a taxonomy automatically. The third company mentioned is Stratify who also makes a claim on the taxonomy generation and unstructured to structured data spaces. The section on ClearForest provides some interesting tidbits on how Compugen used the system to mine the entire Medline database. It is easy to imagine how similar applications involving the mining of patent abstract databases can be easily made.
Posted: Mon - April 7, 2003 at 08:13 PM   Patinformatics   Interesting Reference Articles   Email Comments


© Anthony Trippe