6
Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

  • View
    215

  • Download
    2

Embed Size (px)

Citation preview

Page 1: Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Anaphora Resolution

Sanghoon KwakTakahiro Aoyama

Page 2: Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Anaphora Resolution

• Moussa said the presidents discussed the situation in Iraq and would return to it Monday.

However, he would not comment on whether Syria and Egypt were planning to send planes to Baghdad.

• Anaphora resolution is the process of determining whether a noun phrase refers to an antecedent noun phrase.

Page 3: Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Approach

• Extracted 12 features for each noun phrase pair from a news article corpora

• Constructed a training set using these 12 features. Used 70 news articles from NY Times and Broadcast News.

• Built a decision tree with this training data using the C4.5 decision tree builder

Page 4: Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Features

• Some of 12 features include… Distance Distance by the number of sentences Pronoun Whether a noun phrase is a pronoun or not String match Whether two phrases have the same character sequence Semantic Class Whether two phrases are in the same semantic class

(Person, Organization, etc.) Number

Whether two phrases are both plural or both singular

Page 5: Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Result

Training Testing

Phrase Pairs 406,400 212,865

Incorrectly Classified 24,762 14,213

Error Rate 6.1% 6.7%

• Training and Test Result

• Rules based on decision tree DIST <= 0, JPRO = True, NUM = True, SEMCLASS = True Coreferent

IPRO = True, STR MATCH = true, DIST > 0 Not Coreferent

Page 6: Anaphora Resolution Sanghoon Kwak Takahiro Aoyama

Future Work

• Train and test data from different domains (Blog posts, E-Mail messages, etc.)

• Further subdivide decision tree classes