Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions

Stefan Ruseti, Mihai Dascalu, Stefan Trausan-Matu, Mihai Masala, Gabriel Gutu, Traian Rebedea

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

    13 Downloads (Pure)

    Abstract

    Multi-participant chat conversations are one of the most frequently employed Computer Supported Collaborative Learning tools due to their ease of use. Moreover, chats enhance knowledge sharing, sustain creativity and aid in collaborative problem solving. Nevertheless, the manual analysis of multi-participant chats is a difficult task due to the mixture of different topics and the inter-twinning of multiple discussion threads during the same conversation. Several tools that employ Natural Language Processing techniques have been developed to automatically identify links between contributions in order to facilitate the tracking of topics and of discussion threads, as well as to highlight key contributions in terms of follow-up impact. This paper proposes a novel method for detecting implicit links based on features computed using string kernels and word embeddings, combined with neural networks. This method significantly outperforms previous results on the same dataset. Due to its smaller size, our model represents an alternative to more complex deep neural networks, especially when limited training data is available as is the case of CSCL chats in a specific domain.
    Original languageEnglish
    Title of host publication13th European Conference on Technology Enhanced Learning (EC-TEL 2018)
    EditorsV. Pammer-Schindler, M. Pérez-Sanagustín, H. Drachsler, R. Elferink, M. Scheffel
    PublisherSpringer UK
    Pages482-496
    ISBN (Electronic)978-3-319-98572-5
    ISBN (Print)978-3-319-98571-8
    DOIs
    Publication statusPublished - 2018
    EventEuropean Conference on Technology Enhanced Learning: Lifelong Technology-Enhanced Learning (EC-TEL 2018) - Leeds, United Kingdom
    Duration: 3 Sep 20186 Sep 2018
    https://link.springer.com/conference/ectel

    Conference

    ConferenceEuropean Conference on Technology Enhanced Learning
    Abbreviated titleEC-TEL 2018
    CountryUnited Kingdom
    CityLeeds
    Period3/09/186/09/18
    Internet address

    Fingerprint

    Twinning
    Neural networks
    Processing
    Deep neural networks

    Keywords

    • Computer Supported Collaborative Learning
    • Implicit links identification
    • Natural Language Processing
    • String kernels
    • Neural networks

    Cite this

    Ruseti, S., Dascalu, M., Trausan-Matu, S., Masala, M., Gutu, G., & Rebedea, T. (2018). Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions. In V. Pammer-Schindler, M. Pérez-Sanagustín, H. Drachsler, R. Elferink, & M. Scheffel (Eds.), 13th European Conference on Technology Enhanced Learning (EC-TEL 2018) (pp. 482-496). Springer UK. https://doi.org/10.1007/978-3-319-98572-5_37
    Ruseti, Stefan ; Dascalu, Mihai ; Trausan-Matu, Stefan ; Masala, Mihai ; Gutu, Gabriel ; Rebedea, Traian. / Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions. 13th European Conference on Technology Enhanced Learning (EC-TEL 2018). editor / V. Pammer-Schindler ; M. Pérez-Sanagustín ; H. Drachsler ; R. Elferink ; M. Scheffel. Springer UK, 2018. pp. 482-496
    @inproceedings{67bdffa677464a7da4d5a77a051a58cd,
    title = "Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions",
    abstract = "Multi-participant chat conversations are one of the most frequently employed Computer Supported Collaborative Learning tools due to their ease of use. Moreover, chats enhance knowledge sharing, sustain creativity and aid in collaborative problem solving. Nevertheless, the manual analysis of multi-participant chats is a difficult task due to the mixture of different topics and the inter-twinning of multiple discussion threads during the same conversation. Several tools that employ Natural Language Processing techniques have been developed to automatically identify links between contributions in order to facilitate the tracking of topics and of discussion threads, as well as to highlight key contributions in terms of follow-up impact. This paper proposes a novel method for detecting implicit links based on features computed using string kernels and word embeddings, combined with neural networks. This method significantly outperforms previous results on the same dataset. Due to its smaller size, our model represents an alternative to more complex deep neural networks, especially when limited training data is available as is the case of CSCL chats in a specific domain.",
    keywords = "Computer Supported Collaborative Learning, Implicit links identification, Natural Language Processing, String kernels, Neural networks",
    author = "Stefan Ruseti and Mihai Dascalu and Stefan Trausan-Matu and Mihai Masala and Gabriel Gutu and Traian Rebedea",
    note = "This publication reflects only the author's view and the European Commission is not responsible for any use that may be made of the information it contains.",
    year = "2018",
    doi = "10.1007/978-3-319-98572-5_37",
    language = "English",
    isbn = "978-3-319-98571-8",
    pages = "482--496",
    editor = "V. Pammer-Schindler and M. P{\'e}rez-Sanagust{\'i}n and H. Drachsler and R. Elferink and M. Scheffel",
    booktitle = "13th European Conference on Technology Enhanced Learning (EC-TEL 2018)",
    publisher = "Springer UK",
    address = "United Kingdom",

    }

    Ruseti, S, Dascalu, M, Trausan-Matu, S, Masala, M, Gutu, G & Rebedea, T 2018, Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions. in V Pammer-Schindler, M Pérez-Sanagustín, H Drachsler, R Elferink & M Scheffel (eds), 13th European Conference on Technology Enhanced Learning (EC-TEL 2018). Springer UK, pp. 482-496, European Conference on Technology Enhanced Learning, Leeds, United Kingdom, 3/09/18. https://doi.org/10.1007/978-3-319-98572-5_37

    Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions. / Ruseti, Stefan; Dascalu, Mihai; Trausan-Matu, Stefan; Masala, Mihai; Gutu, Gabriel; Rebedea, Traian.

    13th European Conference on Technology Enhanced Learning (EC-TEL 2018). ed. / V. Pammer-Schindler; M. Pérez-Sanagustín; H. Drachsler; R. Elferink; M. Scheffel. Springer UK, 2018. p. 482-496.

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

    TY - GEN

    T1 - Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions

    AU - Ruseti, Stefan

    AU - Dascalu, Mihai

    AU - Trausan-Matu, Stefan

    AU - Masala, Mihai

    AU - Gutu, Gabriel

    AU - Rebedea, Traian

    N1 - This publication reflects only the author's view and the European Commission is not responsible for any use that may be made of the information it contains.

    PY - 2018

    Y1 - 2018

    N2 - Multi-participant chat conversations are one of the most frequently employed Computer Supported Collaborative Learning tools due to their ease of use. Moreover, chats enhance knowledge sharing, sustain creativity and aid in collaborative problem solving. Nevertheless, the manual analysis of multi-participant chats is a difficult task due to the mixture of different topics and the inter-twinning of multiple discussion threads during the same conversation. Several tools that employ Natural Language Processing techniques have been developed to automatically identify links between contributions in order to facilitate the tracking of topics and of discussion threads, as well as to highlight key contributions in terms of follow-up impact. This paper proposes a novel method for detecting implicit links based on features computed using string kernels and word embeddings, combined with neural networks. This method significantly outperforms previous results on the same dataset. Due to its smaller size, our model represents an alternative to more complex deep neural networks, especially when limited training data is available as is the case of CSCL chats in a specific domain.

    AB - Multi-participant chat conversations are one of the most frequently employed Computer Supported Collaborative Learning tools due to their ease of use. Moreover, chats enhance knowledge sharing, sustain creativity and aid in collaborative problem solving. Nevertheless, the manual analysis of multi-participant chats is a difficult task due to the mixture of different topics and the inter-twinning of multiple discussion threads during the same conversation. Several tools that employ Natural Language Processing techniques have been developed to automatically identify links between contributions in order to facilitate the tracking of topics and of discussion threads, as well as to highlight key contributions in terms of follow-up impact. This paper proposes a novel method for detecting implicit links based on features computed using string kernels and word embeddings, combined with neural networks. This method significantly outperforms previous results on the same dataset. Due to its smaller size, our model represents an alternative to more complex deep neural networks, especially when limited training data is available as is the case of CSCL chats in a specific domain.

    KW - Computer Supported Collaborative Learning

    KW - Implicit links identification

    KW - Natural Language Processing

    KW - String kernels

    KW - Neural networks

    U2 - 10.1007/978-3-319-98572-5_37

    DO - 10.1007/978-3-319-98572-5_37

    M3 - Conference article in proceeding

    SN - 978-3-319-98571-8

    SP - 482

    EP - 496

    BT - 13th European Conference on Technology Enhanced Learning (EC-TEL 2018)

    A2 - Pammer-Schindler, V.

    A2 - Pérez-Sanagustín, M.

    A2 - Drachsler, H.

    A2 - Elferink, R.

    A2 - Scheffel, M.

    PB - Springer UK

    ER -

    Ruseti S, Dascalu M, Trausan-Matu S, Masala M, Gutu G, Rebedea T. Help Me Understand This Conversation: Methods of Identifying Implicit Links Between CSCL Contributions. In Pammer-Schindler V, Pérez-Sanagustín M, Drachsler H, Elferink R, Scheffel M, editors, 13th European Conference on Technology Enhanced Learning (EC-TEL 2018). Springer UK. 2018. p. 482-496 https://doi.org/10.1007/978-3-319-98572-5_37