Artificial Intelligence and Information Retrieval.

ARTIFICIAL intelligence; BUSINESS enterprises; INFORMATION retrieval; INFORMATION scientists; INTERNET searching

This article emphasizes how artificial intelligence can help the businesses build better information retrieval (IR) systems to better serve the needs of knowledge workers. The task of recommending documents to knowledge workers differs from the task of recommending products to consumers. Collaborative filtering, as applied to books, videos, and the like, is a method that attempts to build on patterns of shared tastes or interests among the buying habits of individual shoppers in order to augment conventional search results. Often in information seeking, users search a selected database, not realizing that relevant documents may exist in other databases. Document recommendation can overcome this problem by matching the user's query and other contextual information against indices or profiles of other documents. These ancillary documents may add value to the user's search by digesting or summarizing primary documents or by providing useful encyclopedic or dictionary information. The programmatic identification and tagging of referring expressions in free text is often called information extraction rather than information retrieval, although it is clearly an IR task. One complication in entity extraction is that proper names typically have shortened forms that may become ambiguous when viewed in isolation. Knowledge workers in the 21st century want tools for finding relevant documents, extracting relevant information for them, and assimilating them into existing document classification systems.