Multilingual Support for Lexicon-Based Sentiment Analysis Guided by Semantics.

Alexander Hogenboom, Bas Heerschop, Flavius Frasincar, Uzay Kaymak, Franciska M.G. de Jong

    Research output: Contribution to journalArticleAcademicpeer-review

    43 Citations (Scopus)

    Abstract

    Many sentiment analysis methods rely on sentiment lexicons, containing words and their associated sentiment, and are tailored to one specific language. Yet, the ever-growing amount of data in different languages on the Web renders multi-lingual support increasingly important. In this paper, we assess various methods for supporting an additional target language in lexicon-based sentiment analysis. As a baseline, we automatically translate text into a reference language for which a sentiment lexicon is available, and subsequently analyze the translated text. Second, we consider mapping sentiment scores from a semantically enabled sentiment lexicon in the reference language to a new target sentiment lexicon, by traversing relations between language-specific semantic lexicons. Last, we consider creating a target sentiment lexicon by propagating sentiment of seed words in a semantic lexicon for the target language. When extending sentiment analysis from English to Dutch, mapping sentiment across languages by exploiting relations between semantic lexicons yields a significant performance improvement over the baseline of about 29% in terms of accuracy and macro-level F1 on our data. Propagating sentiment in language-specific semantic lexicons can outperform the baseline by up to about 47%, depending on the seed set of sentiment-carrying words. This indicates that sentiment is not only linked to word meanings, but tends to have a language-specific dimension as well.
    Original languageUndefined
    Pages (from-to)43-53
    Number of pages11
    JournalDecision support systems
    Volume62
    Issue number1
    DOIs
    Publication statusPublished - Jun 2014

    Keywords

    • EWI-24918
    • Machine translation
    • Multi-lingual sentiment analysis
    • Lexicon
    • Propagation
    • Semantics
    • METIS-305950
    • IR-91769
    • Map

    Cite this

    Hogenboom, Alexander ; Heerschop, Bas ; Frasincar, Flavius ; Kaymak, Uzay ; de Jong, Franciska M.G. / Multilingual Support for Lexicon-Based Sentiment Analysis Guided by Semantics. In: Decision support systems. 2014 ; Vol. 62, No. 1. pp. 43-53.
    @article{edc0b4cda8814034a096228afa7a3c33,
    title = "Multilingual Support for Lexicon-Based Sentiment Analysis Guided by Semantics.",
    abstract = "Many sentiment analysis methods rely on sentiment lexicons, containing words and their associated sentiment, and are tailored to one specific language. Yet, the ever-growing amount of data in different languages on the Web renders multi-lingual support increasingly important. In this paper, we assess various methods for supporting an additional target language in lexicon-based sentiment analysis. As a baseline, we automatically translate text into a reference language for which a sentiment lexicon is available, and subsequently analyze the translated text. Second, we consider mapping sentiment scores from a semantically enabled sentiment lexicon in the reference language to a new target sentiment lexicon, by traversing relations between language-specific semantic lexicons. Last, we consider creating a target sentiment lexicon by propagating sentiment of seed words in a semantic lexicon for the target language. When extending sentiment analysis from English to Dutch, mapping sentiment across languages by exploiting relations between semantic lexicons yields a significant performance improvement over the baseline of about 29{\%} in terms of accuracy and macro-level F1 on our data. Propagating sentiment in language-specific semantic lexicons can outperform the baseline by up to about 47{\%}, depending on the seed set of sentiment-carrying words. This indicates that sentiment is not only linked to word meanings, but tends to have a language-specific dimension as well.",
    keywords = "EWI-24918, Machine translation, Multi-lingual sentiment analysis, Lexicon, Propagation, Semantics, METIS-305950, IR-91769, Map",
    author = "Alexander Hogenboom and Bas Heerschop and Flavius Frasincar and Uzay Kaymak and {de Jong}, {Franciska M.G.}",
    note = "eemcs-eprint-24918",
    year = "2014",
    month = "6",
    doi = "10.1016/j.dss.2014.03.004",
    language = "Undefined",
    volume = "62",
    pages = "43--53",
    journal = "Decision support systems",
    issn = "0167-9236",
    publisher = "Elsevier",
    number = "1",

    }

    Multilingual Support for Lexicon-Based Sentiment Analysis Guided by Semantics. / Hogenboom, Alexander; Heerschop, Bas; Frasincar, Flavius; Kaymak, Uzay; de Jong, Franciska M.G.

    In: Decision support systems, Vol. 62, No. 1, 06.2014, p. 43-53.

    Research output: Contribution to journalArticleAcademicpeer-review

    TY - JOUR

    T1 - Multilingual Support for Lexicon-Based Sentiment Analysis Guided by Semantics.

    AU - Hogenboom, Alexander

    AU - Heerschop, Bas

    AU - Frasincar, Flavius

    AU - Kaymak, Uzay

    AU - de Jong, Franciska M.G.

    N1 - eemcs-eprint-24918

    PY - 2014/6

    Y1 - 2014/6

    N2 - Many sentiment analysis methods rely on sentiment lexicons, containing words and their associated sentiment, and are tailored to one specific language. Yet, the ever-growing amount of data in different languages on the Web renders multi-lingual support increasingly important. In this paper, we assess various methods for supporting an additional target language in lexicon-based sentiment analysis. As a baseline, we automatically translate text into a reference language for which a sentiment lexicon is available, and subsequently analyze the translated text. Second, we consider mapping sentiment scores from a semantically enabled sentiment lexicon in the reference language to a new target sentiment lexicon, by traversing relations between language-specific semantic lexicons. Last, we consider creating a target sentiment lexicon by propagating sentiment of seed words in a semantic lexicon for the target language. When extending sentiment analysis from English to Dutch, mapping sentiment across languages by exploiting relations between semantic lexicons yields a significant performance improvement over the baseline of about 29% in terms of accuracy and macro-level F1 on our data. Propagating sentiment in language-specific semantic lexicons can outperform the baseline by up to about 47%, depending on the seed set of sentiment-carrying words. This indicates that sentiment is not only linked to word meanings, but tends to have a language-specific dimension as well.

    AB - Many sentiment analysis methods rely on sentiment lexicons, containing words and their associated sentiment, and are tailored to one specific language. Yet, the ever-growing amount of data in different languages on the Web renders multi-lingual support increasingly important. In this paper, we assess various methods for supporting an additional target language in lexicon-based sentiment analysis. As a baseline, we automatically translate text into a reference language for which a sentiment lexicon is available, and subsequently analyze the translated text. Second, we consider mapping sentiment scores from a semantically enabled sentiment lexicon in the reference language to a new target sentiment lexicon, by traversing relations between language-specific semantic lexicons. Last, we consider creating a target sentiment lexicon by propagating sentiment of seed words in a semantic lexicon for the target language. When extending sentiment analysis from English to Dutch, mapping sentiment across languages by exploiting relations between semantic lexicons yields a significant performance improvement over the baseline of about 29% in terms of accuracy and macro-level F1 on our data. Propagating sentiment in language-specific semantic lexicons can outperform the baseline by up to about 47%, depending on the seed set of sentiment-carrying words. This indicates that sentiment is not only linked to word meanings, but tends to have a language-specific dimension as well.

    KW - EWI-24918

    KW - Machine translation

    KW - Multi-lingual sentiment analysis

    KW - Lexicon

    KW - Propagation

    KW - Semantics

    KW - METIS-305950

    KW - IR-91769

    KW - Map

    U2 - 10.1016/j.dss.2014.03.004

    DO - 10.1016/j.dss.2014.03.004

    M3 - Article

    VL - 62

    SP - 43

    EP - 53

    JO - Decision support systems

    JF - Decision support systems

    SN - 0167-9236

    IS - 1

    ER -