Aspect-based summarization for game review using double propagation
Kevin Yauris, Masayu Leylia Khodra · 2017
Game review consists of opinion on many different aspects. Aspect-based summarization may help to find specific required opinion from a collection of reviews. This paper aims to build aspect based summarization system employing modified Double Propagation (DP) for extracting pairs of aspects and its sentiment, aggregate them by aspect categorization, and presenting them as a summary. DP propagation algorithm is modified to extract target-opinion word pair. We also add additional pruning method for extracted opinion word with DP and add new methods to detect target phrase. Aspect categorization is conducted by defining user-defined taxonomy and calculating similarity score between seed in every aspect category with aspect expression that will be categorized. The summary is presented by grouping aspect expressions of each aspect category based its sentiment orientation. Based on evaluation the proposed summarizer can be applied to game review to present aspect-based summarization that has overall performance F-Measure of 0.5136. The performance still needs some improvements in future works to handle various extraction error.