Aspect-based sentiment analysis is an important task in the fields of affective computing and social computing, which aims to infer the sentiment towards a given aspect. Previous studies have shown notable success when sufficient labeled training data is available. However, annotating adequate data is labor-intensive with high time costs, which sets substantial barriers for generalizing the sentiment predictor to the new domain. Two main challenges exist in cross-domain aspect-based sentiment analysis. One challenge is acquiring the domain-invariant knowledge. As a bridge of domain migration, social text is short, informal and sparse, which brings great challenges to the mining of domain invariant information. The other challenge is mining the syntactic-related words towards the aspect-term. The lack of deep understanding between syntactic-related words and target will affect the recognition performance.

Recently, Prof. Fu Xianghua, deputy dean of the College of Big Data and Internet from Shenzhen Technology University (SZTU), Assistant Prof. Zhang Bowen from the same College and their research team proposed a transformer-based semantic-primary knowledge transferring network (TSPKT) for cross-domain aspect-term sentiment analysis. TSPKT builds an S-Graph from external semantic lexicons, and extract the semantic-primary knowledge from the S-Graph. Moreover, AoaGraphormer is proposed to learn the syntactically relevant words towards the aspect-term. Experiments demonstrate the superiority of TSPKT against the state-of-the-art baseline methods.

Transformer-based semantic-primary knowledge transferring network (TSPKT) [Photo/https://ieeexplore.ieee.org/document/10025370]

The research team published an article titled “Cross-Domain Aspect-based Sentiment Classification by Exploiting Domain-Invariant Semantic-primary Feature” in IEEE Transactions on Affective Computing (Q1, IF: 13.99), which is a top journal in the fields of artificial intelligence and affective computing. Assistant Prof. Zhang Bowen is the first author and Prof. Fu Xianghua the co-corresponding author.

The full text of the article can be found at: https://ieeexplore.ieee.org/document/10025370.

Drafted by Christine（黄昭萱）/ International Cooperation & Student Affairs Office

Revised by International Cooperation & Student Affairs Office

Edited by International Cooperation & Student Affairs Office