This Indexing Software Wrestles With Concepts

One long-standing problem in computer science has been to figure out how to index large amounts of text automatically--especially in a way that involves not just key words but complete ideas, thoughts, and concepts. Indexing journal articles for computerized retrieval, for instance, usually requires specially trained workers, who do the work by hand--a costly and lengthy process.

To continue reading this article you must be a Bloomberg Professional Service Subscriber.