Automatic Aspect-Based Sentiment Analysis (AABSA) from Customer Reviews

Online review platforms provide enormous information for users to evaluate products and services. However, the sheer volume of reviews can create information overload that could increase user search costs and cognitive burden. To reduce information overload, in this paper, we propose an Automatic Aspect-Based Sentiment Analysis (AABSA) model to automatically identify key aspects from Chinese online reviews and conduct aspect-based sentiment analysis. We create a hierarchical structure of hypernyms and hyponyms, apply deep-learning-based representation learning and clustering to identify aspects that are the core content in the reviews, and then calculate the sentiment score of each aspect. To evaluate the performance of the identified aspects, we use an econometric model to estimate the impact of each aspect on product sales. We collaborate with one of Asia’s largest online shopping platforms and employ the model in its product review tagging system to help consumers search for product aspects. Compared with benchmark models, our model is both more effective, because it creates a more comprehensive list of aspects that are indicative of customer needs, and more efficient because it is fully automated without any human labor cost.

[1]  Meng Wang,et al.  Aspect Ranking: Identifying Important Product Aspects from Online Consumer Reviews , 2011, ACL.

[2]  V. Dhar,et al.  Does Chatter Matter? The Impact of User-Generated Content on Music Sales , 2007 .

[3]  Regina Barzilay,et al.  Multiple Aspect Ranking Using the Good Grief Algorithm , 2007, NAACL.

[4]  Tong Zhang,et al.  Supervised and Semi-Supervised Text Categorization using LSTM for Region Embeddings , 2016, ICML.

[5]  Christopher S. G. Khoo,et al.  Aspect-based sentiment analysis of movie reviews on discussion boards , 2010, J. Inf. Sci..

[6]  Bing Liu,et al.  Mining Opinion Features in Customer Reviews , 2004, AAAI.

[7]  Matti A. Kaulio.  Customer, consumer and user involvement in product development: A framework and a review of selected methods , 1998 .

[8]  G. Tellis,et al.  Mining Marketing Meaning from Online Chatter: Strategic Brand Analysis of Big Data Using Latent Dirichlet Allocation , 2014 .

[9]  Huey Yee Lee,et al.  Chinese Sentiment Analysis Using Maximum Entropy , 2011 .

[10]  Adam L. Berger,et al.  A Maximum Entropy Approach to Natural Language Processing , 1996, CL.

[11]  Sasha Blair-Goldensohn,et al.  Building a Sentiment Summarizer for Local Service Reviews , 2008 .

[12]  Kiyoaki Shirai,et al.  PhraseRNN: Phrase Recursive Neural Network for Aspect-based Sentiment Analysis , 2015, EMNLP.

[13]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[14]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[15]  Meng Wang,et al.  Product Aspect Ranking and Its Applications , 2014, IEEE Transactions on Knowledge and Data Engineering.

[16]  Yiming Yang,et al.  XLNet: Generalized Autoregressive Pretraining for Language Understanding , 2019, NeurIPS.

[17]  Xiaokui Xiao,et al.  Recursive Neural Conditional Random Fields for Aspect-based Sentiment Analysis , 2016, EMNLP.

[18]  Diederik P. Kingma,et al.  GPU Kernels for Block-Sparse Weights , 2017 .

[19]  Taher H. Haveliwala.  Efficient Computation of PageRank , 1999 .

[20]  Jacob Goldenberg,et al.  Mine Your Own Business: Market-Structure Surveillance Through Text Mining , 2012, Mark. Sci..

[21]  Adiwijaya,et al.  Aspect-based sentiment analysis to review products using Naïve Bayes , 2017 .

[22]  Ilya Sutskever,et al.  Learning to Generate Reviews and Discovering Sentiment , 2017, ArXiv.

[23]  Bo Pang,et al.  Thumbs up? Sentiment Classification using Machine Learning Techniques , 2002, EMNLP.

[24]  C. Perry,et al.  A customer‐oriented new service development process , 2002 .

[25]  Yuji Matsumoto,et al.  Extracting Aspect-Evaluation and Aspect-Of Relations in Opinion Mining , 2007, EMNLP.

[26]  Eric T. Bradlow,et al.  Automated Marketing Research Using Online Customer Reviews , 2011 .

[27]  Yu-Chun Wang,et al.  Aspect-category-based sentiment classification with aspect-opinion relation , 2016, 2016 Conference on Technologies and Applications of Artificial Intelligence (TAAI).

[28]  Kazutaka Shimada,et al.  Aspect Identification of Sentiment Sentences Using A Clustering Algorithm , 2011 .

[29]  Jingbo Zhu,et al.  Multi-aspect opinion polling from textual reviews , 2009, CIKM.

[30]  John R. Hauser,et al.  Identifying Customer Needs from User-Generated Content , 2019, Mark. Sci..

[31]  John G. Breslin,et al.  A Hierarchical Model of Reviews for Aspect-based Sentiment Analysis , 2016, EMNLP.

[32]  Ishita Chakraborty,et al.  Attribute Sentiment Scoring with Online Text Reviews: Accounting for Language Structure and Missing Attributes , 2019, SSRN Electronic Journal.

[33]  Sergey Brin,et al.  The Anatomy of a Large-Scale Hypertextual Web Search Engine , 1998, Comput. Networks.

[34]  Rajeev Motwani,et al.  The PageRank Citation Ranking : Bringing Order to the Web , 1999, WWW 1999.

[35]  Nigel Collier,et al.  Sentiment Analysis using Support Vector Machines with Diverse Information Sources , 2004, EMNLP.

[36]  Xiao Liu,et al.  Large-Scale Cross-Category Analysis of Consumer Review Content on Sales Conversion Leveraging Deep Learning , 2017, AAAI Workshops.

[37]  Panagiotis G. Ipeirotis,et al.  Show me the money!: deriving the pricing power of product features by mining consumer reviews , 2007, KDD '07.