SSP-based construction of evaluation-annotated data for fine-grained aspect-based sentiment analysis

TL;DR

This paper introduces EVAD, a Korean annotated dataset for fine-grained aspect-based sentiment analysis in e-commerce, created using SSP and extensive linguistic resources, and demonstrates its effectiveness with KoBERT and KcBERT models.

Contribution

The paper presents a novel Korean evaluation-annotated corpus for ABSA, utilizing SSP and FSTs, enhancing analysis of user opinions with detailed aspect-value extraction.

Findings

KoBERT achieved F1 score of 0.88 on aspect-value recognition.

KcBERT achieved F1 score of 0.90 on aspect-value recognition.

The dataset enables more accurate and detailed ABSA in e-commerce reviews.

Abstract

We report the construction of a Korean evaluation-annotated corpus, hereafter called 'Evaluation Annotated Dataset (EVAD)', and its use in Aspect-Based Sentiment Analysis (ABSA) extended in order to cover e-commerce reviews containing sentiment and non-sentiment linguistic patterns. The annotation process uses Semi-Automatic Symbolic Propagation (SSP). We built extensive linguistic resources formalized as a Finite-State Transducer (FST) to annotate corpora with detailed ABSA components in the fashion e-commerce domain. The ABSA approach is extended, in order to analyze user opinions more accurately and extract more detailed features of targets, by including aspect values in addition to topics and aspects, and by classifying aspectvalue pairs depending whether values are unary, binary, or multiple. For evaluation, the KoBERT and KcBERT models are trained on the annotated dataset, showing robust performances of F1 0.88 and F1 0.90, respectively, on recognition of aspect-value pairs.