Human-Level Performance on Word Analogy Questions by Latent Relational Analysis


February 4, 2005


Peter Turney


National Research Council


This talk introduces Latent Relational Analysis (LRA), a method for measuring relational similarity. LRA has potential applications in many areas, including information extraction, word sense disambiguation, machine translation, and information retrieval. Relational similarity is correspondence between relations, in contrast with attributional similarity, which is correspondence between attributes. When two words have a high degree of attributional similarity, we call them synonyms. When two pairs of words have a high degree of relational similarity, we say that their relations are analogous. For example, the word pair mason/stone is analogous to the pair carpenter/wood; the relations between mason and stone are highly similar to the relations between carpenter and wood. Past work on semantic similarity measures has mainly been concerned with attributional similarity. For instance, Latent Semantic Analysis (LSA) can measure the degree of similarity between two words, but not between two relations. Recently the Vector Space Model (VSM) of information retrieval has been adapted to the task of measuring relational similarity, achieving a score of 47% on a collection of 374 college-level multiple-choice word analogy questions. In the VSM approach, the relation between a pair of words is characterized by a vector of frequencies of predefined patterns in a large corpus. LRA extends the VSM approach in three ways: (1) the patterns are derived automatically from the corpus (they are not predefined), (2) the Singular Value Decomposition (SVD) is used to smooth the frequency data (it is also used this way in LSA), and (3) automatically generated synonyms are used to explore reformulations of the word pairs. LRA achieves 56% on the 374 analogy questions, statistically equivalent to the average human score of 57%. On the related problem of classifying noun-modifier relations, LRA achieves similar gains over the VSM, while using a smaller corpus.


Peter Turney

Dr. Peter Turney is a Senior Research Officer in the Interactive Information Group of the National Research Council. In 1988, he obtained his PhD from the University of Toronto, where he then accepted a Postdoctoral Fellowship. He joined the NRC in 1989, and he has since worked on a variety of projects, all involving applications of machine learning technology. His recent work focuses on machine learning applied to natural language. He is the author or co-author of more than seventy publications, a past editor of Canadian Artificial Intelligence magazine, and a member of the Advisory Board of the Journal of Artificial Intelligence Research.