@inproceedings{zhang-qian-2020-convolution,
    title = "Convolution over Hierarchical Syntactic and Lexical Graphs for Aspect Level Sentiment Analysis",
    author = "Zhang, Mi  and
      Qian, Tieyun",
    editor = "Webber, Bonnie  and
      Cohn, Trevor  and
      He, Yulan  and
      Liu, Yang",
    booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2020.emnlp-main.286/",
    doi = "10.18653/v1/2020.emnlp-main.286",
    pages = "3540--3549",
    abstract = "The state-of-the-art methods in aspect-level sentiment classification have leveraged the graph based models to incorporate the syntactic structure of a sentence. While being effective, these methods ignore the corpus level word co-occurrence information, which reflect the collocations in linguistics like {\textquotedblleft}nothing special{\textquotedblright}. Moreover, they do not distinguish the different types of syntactic dependency, e.g., a nominal subject relation {\textquotedblleft}food-was{\textquotedblright} is treated equally as an adjectival complement relation {\textquotedblleft}was-okay{\textquotedblright} in {\textquotedblleft}food was okay{\textquotedblright}. To tackle the above two limitations, we propose a novel architecture which convolutes over hierarchical syntactic and lexical graphs. Specifically, we employ a global lexical graph to encode the corpus level word co-occurrence information. Moreover, we build a concept hierarchy on both the syntactic and lexical graphs for differentiating various types of dependency relations or lexical word pairs. Finally, we design a bi-level interactive graph convolution network to fully exploit these two graphs. Extensive experiments on five bench- mark datasets show that our method outperforms the state-of-the-art baselines."
}