Understanding Readers: Conducting Sentiment Analysis of Instagram Captions

Ming Zhan, Tu Ruibo, Yu Qin

    Research output: Chapter in Book/Conference proceedingConference contributionScientificpeer-review

    5 Citations (Scopus)


    The advent of media transition highlights the importance of user-generated content on social media. Amongst the methods of analysis of user-generated content, sentiment analysis is widely used. Nevertheless, few studies use sentiment analysis to investigate user-generated content on Instagram in the context of public libraries. Therefore, this study aims to fill this research gap by conducting sentiment analysis of two million captions on Instagram. Supervised machine learning algorithms were employed to create the classifier. Three opinion polarities and six emotions were ultimately identified via these captions. These polarities provide new insights for understanding readers, thus helping libraries to deliver better services.

    Original languageUndefined/Unknown
    Title of host publicationCSAI '18: Proceedings of the 2018 2nd International Conference on Computer Science and Artificial Intelligence
    ISBN (Print)978-1-4503-6606-9
    Publication statusPublished - 2018
    MoE publication typeA4 Article in a conference publication
    Eventconference - 2nd International Conference on Computer Science and Artificial Intelligence
    Duration: 1 Jan 2018 → …


    Period01/01/18 → …

    Cite this