Báo cáo khoa học: "A Supervised Learning Approach to Automatic Synonym Identification based on Distributional Features" ppt

... (Companion Volume), pages 1–6, Columbus, June 2008. c 2008 Association for Computational Linguistics A Supervised Learning Approach to Automatic Synonym Identification based on Distributional Features Masato ... novel approach to synonym iden- tification based on supervised learning and distributional features, which correspond to the commonality of indivi...

Ngày tải lên: 23/03/2014, 17:20

... trained on a subset of the Wall Street Journal, comprising 21 million tokens. The reported F-measure for pronoun resolution is 81.5. However, (Morton, 2000) only attempts to resolve singular pronouns, ... made to agree with both of its subvalues. 4.6 Antecedent Selection After classification, one non-pronominal an- tecedent has to be found for each pronoun. As BoosTexter assigns confide...

Ngày tải lên: 17/03/2014, 06:20

... expres- sions. So the demonstrative pronoun in (A3) has to have access not only to (B2) but also to (A1). 3 Data 3.1 Corpus Our work is based on twenty randomly chosen Switchboard dialogues. Taken together, ... Germany michael.strube|christoph.mueller @eml.villa-bosch.de Abstract We apply a decision tree based approach to pronoun resolution in spoken dialogue. Our system deals...

Ngày tải lên: 23/03/2014, 19:20

... aid in automatic lexical acquisition techniques. Many other applications are imaginable as well, using the gap location to inform intonation, semantics, collocation frequency, etc. The contribution ... in- terpretation of the sentence, and could contribute to a natural language understanding or machine translation application. Since WH dependencies also tend to distort the surface sub...

Ngày tải lên: 24/03/2014, 03:20

... the Association for Computational Linguistics, pages 328–332, Jeju, Republic of Korea, 8-14 July 2012. c 2012 Association for Computational Linguistics A Meta Learning Approach to Grammatical ... corrections that agree with the tagged answer to the total number of the suggested corrections whereas recall is the ratio of the suggested corrections that agree with the tagged answer...

Ngày tải lên: 30/03/2014, 17:20

... machine learning approach to the automatic evaluation of machine translation Simon Corston-Oliver, Michael Gamon and Chris Brockett Microsoft Research One Microsoft Way Redmond WA 98052, USA {simonco, ... the distance between pronouns (nPronDist) is the highest predictor, followed by the number of second person pronouns (n2ndPersPron), the number of function words (nFunctionWords), an...

Ngày tải lên: 31/03/2014, 04:20

... to mark up events and time expressions. In addition to the identification of expressions, w e of- ten need to interpret them, for instance to com- pute the absolute time an expression refers to. ... relations are a widely accepted example of this. A third approach is to reify events, to quantify them existentially, and to connect them to other objects using predicates bas...

Ngày tải lên: 31/03/2014, 20:20

... representing the containment re- lation between the extents, two combination oper- ators (, ) corresponding to “and” and “or” op- erator of the boolean model, and ordering operator (✸) representing ... Japan kmasuda@is.s.u-tokyo.ac.jp Abstract This paper investigates an application of the ranked region algebra to information retrieval from large scale but unannotated documents. We autom...

Ngày tải lên: 31/03/2014, 03:20

... from a full tag dictionary to no dictionary at all. We introduce the use of a new information-theoretic criterion, varia- tion of information (Meilˇa, 2002), which can be used to compare a gold ... make identical predictions. 5.2 Varying dictionary knowledge In unsupervised learning, it is not always reasonable to assume that a large tag dictionary is available. To determine the effe...

Ngày tải lên: 20/02/2014, 12:20

... mapped to the same vowel in a different acoustic context. For example, cluster 19 is mapped to /ae/ followed by stop conso- nants, while cluster 20 corresponds to /ae/ followed by nasal consonants. ... iteratively split the HMM states based on objective functions. This method achieved high performance in a phone recognition task using a label -to- phone transducer trained from some...

Ngày tải lên: 07/03/2014, 18:20