An unsupervised aspect detection model for sentiment analysis of reviews

Ayoub Bagheri, M. Saraee, Franciska M.G. de Jong

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    23 Citations (Scopus)
    748 Downloads (Pure)

    Abstract

    With the rapid growth of user-generated content on the internet, sentiment analysis of online reviews has become a hot research topic recently, but due to variety and wide range of products and services, the supervised and domain-specific models are often not practical. As the number of reviews expands, it is essential to develop an efficient sentiment analysis model that is capable of extracting product aspects and determining the sentiments for aspects. In this paper, we propose an unsupervised model for detecting aspects in reviews. In this model, first a generalized method is proposed to learn multi-word aspects. Second, a set of heuristic rules is employed to take into account the influence of an opinion word on detecting the aspect. Third a new metric based on mutual information and aspect frequency is proposed to score aspects with a new bootstrapping iterative algorithm. The presented bootstrapping algorithm works with an unsupervised seed set. Finally two pruning methods based on the relations between aspects in reviews are presented to remove incorrect aspects. The proposed model does not require labeled training data and can be applicable to other languages or domains. We demonstrate the effectiveness of our model on a collection of product reviews dataset, where it outperforms other techniques.
    Original languageUndefined
    Title of host publicationProceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013
    Place of PublicationLondon
    PublisherSpringer
    Pages140-151
    Number of pages12
    ISBN (Print)978-3-642-38823-1
    DOIs
    Publication statusPublished - Jun 2013

    Publication series

    NameLecture Notes in Computer Science
    PublisherSpringer Verlag
    Volume7934
    ISSN (Print)0302-9743
    ISSN (Electronic)1611-3349

    Keywords

    • HMI-SLT: Speech and Language Technology
    • METIS-297709
    • IR-86480
    • EWI-23462

    Cite this

    Bagheri, A., Saraee, M., & de Jong, F. M. G. (2013). An unsupervised aspect detection model for sentiment analysis of reviews. In Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013 (pp. 140-151). (Lecture Notes in Computer Science; Vol. 7934). London: Springer. https://doi.org/10.1007/978-3-642-38824-8_12
    Bagheri, Ayoub ; Saraee, M. ; de Jong, Franciska M.G. / An unsupervised aspect detection model for sentiment analysis of reviews. Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013. London : Springer, 2013. pp. 140-151 (Lecture Notes in Computer Science).
    @inproceedings{443dc401a7d240e9a415413f2cce8f10,
    title = "An unsupervised aspect detection model for sentiment analysis of reviews",
    abstract = "With the rapid growth of user-generated content on the internet, sentiment analysis of online reviews has become a hot research topic recently, but due to variety and wide range of products and services, the supervised and domain-specific models are often not practical. As the number of reviews expands, it is essential to develop an efficient sentiment analysis model that is capable of extracting product aspects and determining the sentiments for aspects. In this paper, we propose an unsupervised model for detecting aspects in reviews. In this model, first a generalized method is proposed to learn multi-word aspects. Second, a set of heuristic rules is employed to take into account the influence of an opinion word on detecting the aspect. Third a new metric based on mutual information and aspect frequency is proposed to score aspects with a new bootstrapping iterative algorithm. The presented bootstrapping algorithm works with an unsupervised seed set. Finally two pruning methods based on the relations between aspects in reviews are presented to remove incorrect aspects. The proposed model does not require labeled training data and can be applicable to other languages or domains. We demonstrate the effectiveness of our model on a collection of product reviews dataset, where it outperforms other techniques.",
    keywords = "HMI-SLT: Speech and Language Technology, METIS-297709, IR-86480, EWI-23462",
    author = "Ayoub Bagheri and M. Saraee and {de Jong}, {Franciska M.G.}",
    note = "10.1007/978-3-642-38824-8_12",
    year = "2013",
    month = "6",
    doi = "10.1007/978-3-642-38824-8_12",
    language = "Undefined",
    isbn = "978-3-642-38823-1",
    series = "Lecture Notes in Computer Science",
    publisher = "Springer",
    pages = "140--151",
    booktitle = "Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013",

    }

    Bagheri, A, Saraee, M & de Jong, FMG 2013, An unsupervised aspect detection model for sentiment analysis of reviews. in Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013. Lecture Notes in Computer Science, vol. 7934, Springer, London, pp. 140-151. https://doi.org/10.1007/978-3-642-38824-8_12

    An unsupervised aspect detection model for sentiment analysis of reviews. / Bagheri, Ayoub; Saraee, M.; de Jong, Franciska M.G.

    Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013. London : Springer, 2013. p. 140-151 (Lecture Notes in Computer Science; Vol. 7934).

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    TY - GEN

    T1 - An unsupervised aspect detection model for sentiment analysis of reviews

    AU - Bagheri, Ayoub

    AU - Saraee, M.

    AU - de Jong, Franciska M.G.

    N1 - 10.1007/978-3-642-38824-8_12

    PY - 2013/6

    Y1 - 2013/6

    N2 - With the rapid growth of user-generated content on the internet, sentiment analysis of online reviews has become a hot research topic recently, but due to variety and wide range of products and services, the supervised and domain-specific models are often not practical. As the number of reviews expands, it is essential to develop an efficient sentiment analysis model that is capable of extracting product aspects and determining the sentiments for aspects. In this paper, we propose an unsupervised model for detecting aspects in reviews. In this model, first a generalized method is proposed to learn multi-word aspects. Second, a set of heuristic rules is employed to take into account the influence of an opinion word on detecting the aspect. Third a new metric based on mutual information and aspect frequency is proposed to score aspects with a new bootstrapping iterative algorithm. The presented bootstrapping algorithm works with an unsupervised seed set. Finally two pruning methods based on the relations between aspects in reviews are presented to remove incorrect aspects. The proposed model does not require labeled training data and can be applicable to other languages or domains. We demonstrate the effectiveness of our model on a collection of product reviews dataset, where it outperforms other techniques.

    AB - With the rapid growth of user-generated content on the internet, sentiment analysis of online reviews has become a hot research topic recently, but due to variety and wide range of products and services, the supervised and domain-specific models are often not practical. As the number of reviews expands, it is essential to develop an efficient sentiment analysis model that is capable of extracting product aspects and determining the sentiments for aspects. In this paper, we propose an unsupervised model for detecting aspects in reviews. In this model, first a generalized method is proposed to learn multi-word aspects. Second, a set of heuristic rules is employed to take into account the influence of an opinion word on detecting the aspect. Third a new metric based on mutual information and aspect frequency is proposed to score aspects with a new bootstrapping iterative algorithm. The presented bootstrapping algorithm works with an unsupervised seed set. Finally two pruning methods based on the relations between aspects in reviews are presented to remove incorrect aspects. The proposed model does not require labeled training data and can be applicable to other languages or domains. We demonstrate the effectiveness of our model on a collection of product reviews dataset, where it outperforms other techniques.

    KW - HMI-SLT: Speech and Language Technology

    KW - METIS-297709

    KW - IR-86480

    KW - EWI-23462

    U2 - 10.1007/978-3-642-38824-8_12

    DO - 10.1007/978-3-642-38824-8_12

    M3 - Conference contribution

    SN - 978-3-642-38823-1

    T3 - Lecture Notes in Computer Science

    SP - 140

    EP - 151

    BT - Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013

    PB - Springer

    CY - London

    ER -

    Bagheri A, Saraee M, de Jong FMG. An unsupervised aspect detection model for sentiment analysis of reviews. In Proceedings of the 18th International Conference on Applications of Natural Language to Information Systems, NLDB 2013. London: Springer. 2013. p. 140-151. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-642-38824-8_12