Graph-based Word Clustering Applied to Question Answering and Reading Comprehension Tests

This paper describes our participation in the QA4MRE 2011 task, targeted at reading comprehension tests and multiple choice question answering. Our system constructs a co-occurrence graph with words that are common or proper nouns and verbs extracted from each document. 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 significant co-occurrence is applied to the co-occurrence graph. Each of the detected communities are treated as different contexts of a question in the corpus, and these contexts are used to find the most suitable answer. Our evaluation results suggest that, the number of retrieved documents is an important factor in the results.