Aspect-Level Sentiment Analysis Incorporating Location Information and Interaction Attention

doi:10.3778/j.issn.1002-8331.2311-0035

Abstract

Abstract: A large number of commentary texts appear on social media and e-commerce platforms, and attention-based aspect-level sentiment analysis methods have been widely used to analyze these texts.When the existing methods achieve the interactive attention between aspect words and context, there is no use of the relative positional relationship between context and aspect words, and only focus on the influence of aspect words on the context, resulting in insufficient semantic interaction, and the aspect words as a whole. An aspect-level sentiment analysis model with interactive attention incorporating relative position information is proposed. Firstly, the bidirectional long short-term memory network is used to learn the semantic features of the aspect words and context that fuses the location information, and then the learnable parameter matrix is integrated to conduct interactive learning of the semantic features of context and aspect words,and interactive attention is used to calculate the impact of aspect words on context and the impact of context on aspect words at word granularity. Finally, sentiment classification is carried out. Several experiments are conducted on the SemEval 2014 Task4 and Twitter benchmark datasets. The experimental results show that the proposed model achieves better performance than the comparison methods.

Key words: aspect-based sentiment analysis, position information, interactive attention, deep learning

摘要： 社交媒体和电商平台中涌现了大量的评论性文本，基于注意力的方面级情感分析方法已经被广泛用于对这些文本进行分析。现有的方法在实现方面词和上下文的交互注意时，存在没有利用上下文和方面词的相对位置关系，只关注方面词对上下文的影响导致语义交互不够充分，和将方面词作为一个整体进行计算等问题。提出一种融合相对位置信息的交互注意力的方面级情感分析模型。利用双向长短期记忆网络学习融合位置信息的上下文和方面词的语义特征，融入可学习的参数矩阵将上下文与方面词的语义特征进行交互学习，并使用交互注意力在字词粒度上分别计算方面词对上下文的影响以及上下文对方面词的影响，最后进行情感分类。在SemEval 2014 Task4基准数据集以及Twitter数据集上进行了多个实验，实验结果表明，提出的模型取得的性能优于比较方法。

关键词: 方面级情感分析, 位置信息, 交互注意力, 深度学习

LI Jiajing, LI Sheng, DAI Yuanyuan, MENG Tao, LUO Xiaoqing, YAN Hongfei. Aspect-Level Sentiment Analysis Incorporating Location Information and Interaction Attention[J]. Computer Engineering and Applications, 2025, 61(6): 220-228.

李佳静, 李盛, 戴媛媛, 孟涛, 罗小清, 闫宏飞. 融合位置信息和交互注意力的方面级情感分析[J]. 计算机工程与应用, 2025, 61(6): 220-228.

References

[1] NGUYEN H T, NGUYEN L M. ILWAANet: an interactive lexicon-aware word-aspect attention network for aspect-level sentiment classification on social networking[J]. Expert Systems with Application, 2020, 146: 113065.
[2] PARK H J, SONG M, SHIN K S. Deep learning models and datasets for aspect term sentiment classification: implementing holistic recurrent attention on target-dependent memories[J]. Knowledge-Based Systems, 2020, 187: 104825.
[3] NAZIR A, RAO Y, WU L, et al. Issues and challenges of aspect-based sentiment analysis: a comprehensive survey[J]. IEEE Transactions on Affective Computing, 2021, 226：107134.
[4] WANG J, JIE L, LI S, et al. Aspect sentiment classification with both word-level and clause-level attention networks[C]//Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI-18), 2018.
[5] 王婷, 杨文忠. 文本情感分析方法研究综述[J]. 计算机工程与应用, 2021, 57(12): 11-24.
WANG?T, YANG?W Z. Review of? text sentiment? analysis methods[J]. Computer Engineering and Applications, 2021, 57(12): 11-24.
[6] KRISHNA A, AKHILESH V, AICH A, et al. Sentiment analysis of restaurant reviews using machine learning techniques[C]//Proceedings of International Conference on Emerging Research in Electronics, Computer Science and Technology, (ICERECT 2018), 2019: 687-696.
[7] AWWALU J, BAKAR A A, YAAKUB M R. Hybrid N-gram model using Na?ve Bayes for classification of political sentiments on Twitter[J]. Neural Computing and Applications, 2019, 31: 9207-9220.
[8] KALCHBRENNER N, GREFENSTETTE E, BLUNSOM P. A convolutional neural network for modelling sentences[J]. arXiv:1404.2188, 2014.
[9] WANG Y, HUANG M, ZHU X, et al. Attention-based LSTM for aspect-level sentiment classification[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, 2016: 606-615.
[10] MA D, LI S, ZHANG X, et al. Interactive attention networks for aspect-level sentiment classification[J]. arXiv:1709.00893, 2017.
[11] HUANG B, OU Y, CARLEY K M. Aspect level sentiment classification with attention-over-attention neural networks[C]//Proceedings of the 11th International Conference on Social, Cultural, and Behavioral Modeling (SBP-BRiMS 2018), Washington, DC, USA, 2018: 197-206.
[12] GU S, ZHANG L, HOU Y, et al. A position-aware bidirectional attention network for aspect-level sentiment analysis[C]//Proceedings of the 27th International Conference on Computational Linguistics, 2018: 774-784.
[13] PONTIKI M, GALANIS D, PAVLOPOULOS J, et al. SemEval-2014 task 4: aspect based sentiment analysis[C]//Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), 2014.
[14] DONG L, WEI F, TAN C, et al. Adaptive recursive neural network for target-dependent twitter sentiment classification[C]//Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), 2014: 49-54.
[15] TANG D, QIN B, LIU T. Aspect level sentiment classification with deep memory network[J]. arXiv:1605.08900, 2016.
[16] SONG Y, WANG J, JIANG T, et al. Attentional encoder network for targeted sentiment classification[J]. arXiv:1902. 09314, 2019.
[17] SCARIA K, GUPTA H, SAWANT S A, et al. InstructABSA: instruction learning for aspect based sentiment analysis[J]. arXiv:2302.08624, 2023.
[18] ZHANG C, LI Q, SONG D. Aspect-based sentiment classification with aspect-specific graph convolutional networks[J]. arXiv:1909.03477, 2019.
[19] ZHU Z, ZHANG D, LI L, et al. Knowledge-guided multi-granularity GCN for ABSA[J]. Information Processing & Management, 2023, 60(2): 103223.
[20] PENNINGTON J, SOCHER R, MANNING C D. Glove: global vectors for word representation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2014: 1532-1543.
[21] DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[J]. arXiv:1810.04805, 2018.
[22] SCHUSTER M, PALIWAL K K. Bidirectional recurrent neural networks[J]. IEEE Transactions on Signal Processing, 1997, 45(11): 2673-2681.
[23] TANG D, QIN B, FENG X, et al. Effective LSTMs for target-dependent sentiment classification[J]. arXiv:1512.01100, 2015.
[24] ZHOU J, HUANG J X, HU Q V, et al. SK-GCN: modeling syntax and knowledge via graph convolutional network for aspect-level sentiment classification[J]. Knowledge-Based Systems, 2020, 205(3): 106292.
[25] XU?H, LIU?B, SHU?L, et?al. BERT?post-training?for?review reading?comprehension?and?aspect-based?sentiment?analysis[J]. arXiv:1904.02232, 2019.
[26] CHEN F, YUAN Z, HUANG Y. Multi-source data fusion for aspect-level sentiment classification[J]. Knowledge-Based Systems, 2019, 187: 104831.
[27] TIAN Y, CHEN G, SONG Y. Enhancing aspect-level sentiment analysis with word dependencies[C]//Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics, 2021: 3726-3739.
[28] ZHAO M, YANG J, ZHANG J, et al. Aggregated graph convolutional networks for aspect-based sentiment classification[J]. Information Sciences, 2022, 600: 73-93.
[29] RADFORD?A, NARASIMHAN?K, SALIMANS?T, et?al. Improving?language?understanding?by?generative?pre?training[J]. arXiv:1801.06146. 2018.