A challenge for building the Semantic Web is to capture the knowledge contained in large document collections without having to perform massive manual knowledge engineering. This paper presents Breeder, a system designed to semi-automatically generate lists of “similar” items based on their contexts in a large corpus. To breed a single item list, the system uses a small set of user-provided seed items as the starting point in a bootstrapping operation which can either proceed automatically until a stopping condition is met or be guided by interactive user feedback as the bootstrapping proceeds. The bootstrapping relies on the “duality” of two orthogonal views of the items and their features, as obtained by text analysis. We also present MultiClassBreeder, which further constrains breeding of the classes as they compete for members with one another.
We have experimented using Breeder to build lexical classes of nominal expressions, a result which is useful for populating a knowledge base with instances of ontological classes. We will describe Breeder’s implementation, including the underlying text analysis and measures taken to make real-time, interactive use feasible. We will discuss our preliminary evaluation of Breeder. We also describe our plans to use Breeder to generate pairs of nominal expressions, referring to class instances, which can be used to populate knowledge bases with the extensions of ontological relations.
By: Roy J. Byrd
Published in: RC22967 in 2003
LIMITED DISTRIBUTION NOTICE:
This Research Report is available. This report has been submitted for publication outside of IBM and will probably be copyrighted if accepted for publication. It has been issued as a Research Report for early dissemination of its contents. In view of the transfer of copyright to the outside publisher, its distribution outside of IBM prior to publication should be limited to peer communications and specific requests. After outside publication, requests should be filled only by reprints or legally obtained copies of the article (e.g., payment of royalties). I have read and understand this notice and am a member of the scientific community outside or inside of IBM seeking a single copy only.
Questions about this service can be mailed to reports@us.ibm.com .