Combining Independent Modules in Lexical Multiple-Choice Problems

Simple item page
dc.creatorTurney, Peter D.
dc.creatorLittman, Michael L.
dc.creatorBigham, Jeffrey
dc.creatorShnayder, Victor
dc.date2005-01-10
dc.date.accessioned2026-07-25T16:52:18Z
dc.descriptionExisting statistical approaches to natural language problems are very coarse approximations to the true complexity of language processing. As such, no single technique will be best for all problem instances. Many researchers are examining ensemble methods that combine the output of multiple modules to create more accurate solutions. This paper examines three merging rules for combining probability distributions: the familiar mixture rule, the logarithmic rule, and a novel product rule. These rules were applied with state-of-the-art results to two problems used to assess human mastery of lexical semantics -- synonym questions and analogy questions. All three merging rules result in ensembles that are more accurate than any of their component modules. The differences among the three rules are not statistically significant, but it is suggestive that the popular mixture rule is not the best rule for either of the two problems.
dc.description10 pages, related work available at http://www.cs.rutgers.edu/~mlittman/ and http://purl.org/peter.turney/
dc.identifierhttps://arxiv.org/abs/cs/0501018
dc.identifierhttp://arxiv.org/abs/cs/0501018
dc.identifierRecent Advances in Natural Language Processing III: Selected Papers from RANLP 2003, Eds: N. Nicolov, K. Botcheva, G. Angelova, and R. Mitkov, (2004), Current Issues in Linguistic Theory (CILT), 260, John Benjamins, 101-110
dc.identifier.urihttps://dspace.dare.co.zw/handle/123456789/43760
dc.subjectMachine Learning
dc.subjectComputation and Language
dc.subjectInformation Retrieval
dc.subjectI.2.6; I.2.7; H.3.1; J.5
dc.titleCombining Independent Modules in Lexical Multiple-Choice Problems
dc.typetext

Files

Collections

arXiv