Graph-based Word Clustering Applied to Question Answering and Reading Comprehension Tests
Juan Martínez-Romo, Lourdes Araujo · 2015
Abstract. This paper describes our participation in the QA4MRE 2011 task, targeted at reading comprehension tests and multiple choice ques-tion answering. Our system constructs a co-occurrence graph with words that are common or proper nouns and verbs extracted from each docu-ment. The documents are pre-selected through an information retrieval process for recovering only those that are most relevant to a particular question. An algorithm to detect communities of words with signicant co-occurrence is applied to the co-occurrence graph. Each of the detected communities are treated as dierent contexts of a question in the cor-pus, and these contexts are used to nd the most suitable answer. Our evaluation results suggest that, the number of retrieved documents is an important factor in the results.