Automated rule selection for opinion target extraction

Opinion target extraction, also called aspect extraction, aims to extract fine-grained opinion targets from opinion texts, such as customer reviews of products and services. This task is important because opinions without targets are of limited use. It is one of the core tasks of the popular aspect-oriented opinion mining, and is also among the most challenging tasks tackled by opinion mining researchers. Previous work has shown that the syntactic-based approach, which employs extraction rules about grammar dependency relations between opinion words and aspects (or targets), performs quite well. This approach is highly desirable in practice because it is unsupervised and domain independent. The problem of this approach is that the extraction rules should be carefully selected and tuned manually so as not to produce too many errors. Although it is easy to evaluate the accuracy of each rule automatically, it is not easy to select a set of rules that produces the best overall result due to the overlapping coverage of the rules. In this paper, we propose two approaches to select an effective set of rules. The first approach employs a greedy algorithm, and the second one employs a local search algorithm, specifically, simulated annealing. Our experiment results show that the proposed approaches can select a subset of a given rule set to achieve significantly better results than the full rule set and the existing state-of-the-art CRF-based supervised method.