PerSent -- Persian Sentiment Analysis and Opinion Mining Lexicon

This corpus was presented in the paper(s):

Kia Dashtipour, Amir Hussain, Qiang Zhou, Alexander Gelbukh. PerSent: A Freely Available Persian Sentiment Lexicon. In BICS 2016, 8th International Conference on Brain-Inspired Cognitive Systems, 28-30 November 2016, Beijing, China.

Any work that uses the dataset or software should cite the abovementioned paper(s).

License: free for non-commercial academic purposes. Any publication that benefited from these data or software must state the origin of the data and software and cite the abovementioned paper(s). We will be grateful to you if you let us know of the use of the data or software and of citing our papers. Any derived work should specify the original source and its authors and contain this license, including the publication references mentioned above. If you modify this corpus or software, correct errors in it, or add annotation/functionality to it, we will be grateful if you send us the new version, to be available from this site. See also individual license files below.

Persian Sentiment Analysis and Opinion Mining Lexicon

This dataset presents real-valued polarity labels, in the range from -1 to 1, for thousands of Persian words and expressions.

This resource is useful for sentiment analysis, opinion mining, sentiment polarity detection, social network analysis, emotion analysis, etc., in Persian (Farsi) language.

Current version:

PerSent 2.0:

    The second version of the lexicon includes about 1500 Persian words from the first version plus 700 idiomatic words and expressions. The idiom list has been shown particularly useful for analyzing highly informal texts such as user-contributed contents, e.g., movie reviews or product reviews.

    A paper presenting this lexicon is under evaluation. The lexicon, version 2.0, will be publicly available here upon acceptance of the paper.

Previous versions:

PerSent 1.0:

    The first version of the lexicon includes about 1500 Persian words.

    Download the Persian sentiment analysis and opinion mining lexicon for polarity detection, version 1.0, as an Excel file (the dataset is in its second sheet).