Implicit Aspect-Based Opinion Mining and Analysis of Airline Industry Based on User-Generated Reviews

Abstract Mining opinions from reviews has been a field of ever-growing research. These include mining on document level, sentence level and even aspect level. While explicitly mentioned aspects user-generated texts have widely researched, very little work done in gathering that are implied not mentioned. Previous to identify implicit opinion was limited syntactic-based classifiers or other machine learning methods trained restaurant dataset. In this paper, the present is novel study for extracting analysing airline English. Through study, an domain-specific aspect-based annotated corpus, two-way technique first augments pre-trained word embeddings sequential with stochastic gradient descent optimized conditional random fields (CRF) second using ensemble algorithms classify devised developed. This resolves double-implicit problem, most encountered by previous text mining. Experiments hold-out test set i.e., entity extraction CRF yield result ROC-AUC score 96% F 1 94% outperforming few baseline systems. Further experiments range classifier each yields ranging 71 94.8% all entities. two-level classification outperforms many systems domain.