Chinese pronominal anaphora resolution using lexical knowledge and entropy‐based weight

Abstract

Pronominal anaphors are commonly observed in written texts. In this article, effective Chinese pronominal anaphora resolution is addressed by using lexical knowledge acquisition and salience measurement. The lexical knowledge acquisition is aimed to extract more semantic features, such as gender, number, and collocate compatibility by employing multiple resources. The presented salience measurement is based on entropybased weighting on selecting antecedent candidates.The resolution is justified with a real corpus and compared with a rule-based model. Experimental results by five-fold cross-validation show that our approach yields 82.5% success rate on 1343 anaphoric instances. In comparison with a general rule-based approach, the performance is improved by 7%.

When “He” Can Also Be “She”: An ERP Study of Reflexive Pronoun Resolution in Written Mandarin Chinese

“…Third, from an empirical point of view, the asymmetry on gender specificity presented by the pronouns is reflected in the usage of / he as a default. Wu and Liang ( 2008 ) analyzed 150 news items taken from the Academia Sinica Balanced Corpus (ASBC), for a rule-based corpus analysis of Chinese pronominal anaphor resolution. The results showed an error rate of 0.21 for gender mismatches between / he and female antecedents.…”

“…Based on previous ERP studies and the two-stage model of anaphor resolution (Garrod and Sanford, 1994 ), interactions among antecedent gender explicitness, reflexive pronoun gender specificity, and gender congruity are thus expected mainly in the N400 or P600 time intervals. Since the pronoun /he could be considered as a default pronoun (that can refer to both male and female antecedents) (Wu and Liang, 2008 ) and supported by the word frequencies of both pronouns in the Sinica Corpus, two hypotheses are possible about how the default pronoun is perceived for gender. First, if the /he is recognized as male-biased, a gender mismatch N400 or P600 effect is predicted for /himself .…”

Improving Definite Anaphora Resolution by Effective Weight Learning and Web-Based Knowledge Acquisition

“…In our previous research, the distribution ratios of these three types are 57%, 29%, and 14%, respectively. In Chinese zero and pronominal anaphora resolution, it is showed that applying weight learning can significantly improve resolution performance in comparison with un-weighted methods [19]. Therefore, we address the problem of Chinese definite anaphora for overall comprehension of Chinese anaphora resolution tasks.…”