Extraction of Key Expressions Indicating the Important Sentence from Article Abstracts

In this study, we aim to extract key expressions that indicate the important sentence describing the originalities or contributions from article abstracts. The expense of searching academic information increases because of increases in the number of articles, discipline subdivisions, and promotion of interdisciplinary research. Improving the extraction and presentation of the main points from article abstracts will contribute to reducing academic information search expenses. We extracted pseudo-important sentences from each article abstract based on the ratio of the number of words in the identified sentence that appear in the article title to the number of all words in the sentence. After that, we evaluated the ratio of the number of the pseudo-important sentences including each N-gram to the number of all sentences including that N-gram. We then extracted the N-grams with a ratio as high as the key expressions.