Search

In this study, we propose to use two corpus-driven linguistic approaches for a sense prediction study. We will concentrate on the character similarity clustering approach and the concept similarity clustering approach to predict the senses of non-assigned words by using corpora and tools, such as the Chinese Gigaword Corpus and HowNet. In this study, we will evaluate sense predictions via the sense divisions of Chinese Wordnet (CWN) and Xiandai Hanyu Cidian (Xian Han). Using these corpora, we will determine the clusters of our four target words — chi1 "eat", wan2 "play", huan4 "change", and shao1 "burn" — in order to predict all possible senses and then evaluate them. This process will demonstrate the viability of the corpus-based approaches.

MARVS is a theory for the representation of verbal semantics. This paper examines two types of modules and two sets of attributes, namely, role modules, event modules, role-internal attributes, and event-internal attributes, to account for the eventive and semantic factors of verbal polysemy. This study focuses on the polysemous verb chī. Based on MARVS, chī is divided into six senses. These senses are derived metonymically or metaphorically from the prototypical sense.

In this paper, we focus on the reliability of information encoded in a Web 2.0 community platform. Specifically, we aim to explore how linguistically encoded clues can contribute to the task. Given that evidentiality is the linguistic representation for the reliability of a statement, we propose to use evidential, the lexicalized evidentiality, to model text representation under the framework of machine learning based text classification. Based on the model, we conducted experiments to identify the best answers in Collaborative Question Answering (CQA) services. The experimental results confirm that, incorporating evidential for predicting text reliability is effective, since it shows the writer's self-judgment for information reliability. Moreover, our method can largely reduce the dimensionality of the vector space, and therefore provide an improvement in efficiency.

One unique challenge in Chinese Language Processing is cross-strait named entity recognition. Due to the adoption of different transliteration strategies, foreign name transliterations can vary greatly between the PRC and Taiwan, creating difficulties in NLP tasks including data mining, translation and information retrieval. In this paper, we introduce a novel approach to automatic extraction of divergent transliterations of foreign named entities that bootstraps co-occurrence statistics from tagged Chinese corpora, thereby producing higher precision.

Event semantics in general and event type coercion in particular have been a challenging yet rewarding topic in verbal semantics [1]. However, there have been few corpus-based empirical accounts discussing the range of event type coercions based on the lexical meanings of the verbs. In this paper, we explore the possible types of event coercions for two verbs of ingestion in Mandarin Chinese. In particular, we will show that the different types of coercions can be predicted by the bifurcation of event-internal and role-internal attributes proposed in the MARVS theory [2]. Data examined are taken from the Chinese Gigaword Corpus [3, 4] and accessed through Chinese Word Sketch [5].

This study examines whether Mapping Principles, which govern how concrete concepts in a source domain are mapped to abstract concepts in a target domain, are realized in large-scale corpus data. In particular, we propose a frequency-based collocational approach to determine mapping principles based on the most productive mapping. In addition, in contrast to the target-domain-oriented method used in previous studies, we employ the source-domain-oriented method to search for lexical mappings from a particular source domain, i.e., BUILDING, to a number of target domains. The corpus data demonstrate that mapping principles exist in a source-target domain pairing. In addition, we found that different target domains select the source domain of BUILDING for different underlying reasons. Our study follows the lexical analysis of conceptual metaphors proposed by the Conceptual Mapping Model and helps us better understand how semantic networks of lexical words are represented in the lexicon via conceptual mappings.