Mining Users' Intentions from Thai Tweets Using BERT Models


Nattapong Sanchan

    Abstract:

    In this paper, we explore the mining of users’ intentions in text. We viewed that being able to identify the intentions of users expressed in textual data provides us to specifically know aims and what users want to do. In the experiment, we collected tweets, constructed a Thai intention corpus, and performed a binary classification task on the corpus. We investigated the intent classification results derived through the application of 3 different Bidirectional Encoder Representations from Transformers (BERT), Word Embedding, and Bag of Words models. The results revealed that BERT Based EN-TH Cased model outperforms other models in both classification and processing time aspects. It achieves the F1 Score of 0.81 and performs the classification task faster than other BERT models up to 15%.

    Keywords: Intent Mining, Intention Mining, Intent Classification, Intent Detection, Text Mining, Natural Language Processing

    References:

    Download this paper: [PDF]
    Dowload bibtex entry: [BibTex] 
    @article{2023_Sanchan,
        title={Mining Users' Intentions from Thai Tweets Using BERT Models},
        author={Sanchan, Nattapong},
        journal={Journal of Information Science and Technology},
        volume={13},
        number={1},
        pages={17--25},
        year={2023}
      }
    Rich text bibliography entry (for copy & paste into a word processor):
    Sanchan, N. (2023). Mining Users' Intentions from Thai Tweets Using BERT Models. Journal of Information Science and Technology, 13(1), 17-25.