Recent Advances in HNC’s Context Vector Information Retrieval Technology

RECENT ADVANCES IN HNC'S CONTEXT VECTOR INFORMATION RETRIEVAL TECHNOLOGY Marc R. Ilgen, David A. Rushall HNC Software, Inc., 5930 Cornerstone Court West, San Diego, CA 92121 USA emaih mri@hne.eom, dar@hnc.eom ABSTRACT Over the past few years, HNC has developed a neural network based, vector space approach to text retrieval. This approach, embodied in a system called MatchPlus, allows the user to retrieve information on the basis of meaning and context of a fi'ee text query. The MatchPlus system uses a neural network based, constrained sel~organization technique to learn word stem interrelationships directly 3~om a training corpus, thereby eliminating the need for hand crafted linguistic knowledge bases and their often substantial maintenance requirements. This paper presents results fi'om recent enhancements to the basic MatchPlus concept. These enhancements include the development of a one step learning law that greatly reduces the amount of time and~or computational resources required to train the system, and the development of a prototype multilingual (English and Spanish) text retrieval system. 1. INTRODUCTION While the current MatchPlus learning law has proven to be effective in encoding relationships between words, it is computationally intensive and requires multiple passes through the training corpus. The purpose of the one step learning law is to approximate the behavior of the original learning law while performing only a single pass through the training corpus. The one step learning law uses a single pass through the training corpus to obtain desired dot product values for the set of trained context vectors. The desired dot product values are determined on the basis of information theoretic statistical relationships between co-occurring word stems found in the training corpus. Desired dot products are found such that words that tend to co- occur will have context vectors that point in similar directions while words that do not co-occur will have context vectors that tend to be orthogonal. These desired dot products are used to perform a quasi-linear transformation on an initial set of quasi-orthogonal, high dimensional vectors. This vector transformation and subsequent renormalization results in a set of context vectors that represents the relationships between word stems in a near-optimal fashion. The time requirements for training this set of vectors scale as O(Nn) where N is the number of word stems in the vocabulary and n is the average number of word stems found to co-occur (and/or be related to) any given word stem (usually on the order of several hundred). This new learning law reduces the training time by a factor on the order of 100 over the original context vector learning law with little or no degradation in performance. Results can be improved even further by adjusting the dimension of the context vectors. HNC has also developed an approach to learning stem-level relationships across multiple languages and has used this approach to develop a prototype multilingual retrieval system. This technique, called "symmetric learning", is based upon the use of tie words, which provide connectivity between each language's portion of the context vector space. In the symmetric approach, …

Content truncated to 500 words