Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks

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

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

    Abstract

    Chat conversations between more than two participants are often used in Computer Supported Collaborative Learning (CSCL) scenarios because they enhance collaborative knowledge sharing and sustain creativity. However, multi-participant chats are more difficult to follow and analyze due to the complex ways in which different discussion threads and topics can interact. This paper introduces a novel method based on neural networks for detecting implicit links that uses features computed with string kernels and word embeddings. In contrast to previous experiments with an accuracy of 33%, we obtained a considerable increase to 44% for the same dataset. Our method represents an alternative to more complex deep neural networks that cannot be properly used due to overfitting on limited training data.
    Original languageEnglish
    Title of host publicationArtificial Intelligence in Education
    Subtitle of host publication19th International Conference, AIED 2018, London, UK, June 27–30, 2018, Proceedings, Part II
    EditorsC.P. Rosé, R. Martínez-Maldonado, H.U. Hoppe, R. Luckin, M. Mavrikis, K. Porayska-Pomsta, B. McLaren, B. d. Boulay
    Place of PublicationCham
    PublisherSpringer
    Pages204-208
    Number of pages5
    Edition1
    ISBN (Electronic)9783319938462
    ISBN (Print)9783319938455
    DOIs
    Publication statusPublished - Jun 2018
    EventArtificial Intelligence in Education: 19th International Conference, AIED 2018, London, UK, June 27 – 20, 2018 - London, United Kingdom
    Duration: 27 Jun 201830 Jun 2018
    https://aied2018.utscic.edu.au/

    Publication series

    NameLecture Notes in Computer Science
    PublisherSpringer

    Conference

    ConferenceArtificial Intelligence in Education
    Abbreviated titleAIED 2018
    CountryUnited Kingdom
    CityLondon
    Period27/06/1830/06/18
    Internet address

    Fingerprint

    Neural networks
    Experiments
    Deep neural networks

    Keywords

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

    Cite this

    Masala, M., Ruseti, S., Gutu-Robu, G., Rebedea, T., Dascalu, M., & Trausan-Matu, S. (2018). Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks. In C. P. Rosé, R. Martínez-Maldonado, H. U. Hoppe, R. Luckin, M. Mavrikis, K. Porayska-Pomsta, B. McLaren, ... B. D. Boulay (Eds.), Artificial Intelligence in Education: 19th International Conference, AIED 2018, London, UK, June 27–30, 2018, Proceedings, Part II (1 ed., pp. 204-208). (Lecture Notes in Computer Science). Cham: Springer. https://doi.org/10.1007/978-3-319-93846-2_37
    Masala, Mihai ; Ruseti, Stefan ; Gutu-Robu, Gabriel ; Rebedea, Traian ; Dascalu, Mihai ; Trausan-Matu, Stefan. / Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks. Artificial Intelligence in Education: 19th International Conference, AIED 2018, London, UK, June 27–30, 2018, Proceedings, Part II. editor / C.P. Rosé ; R. Martínez-Maldonado ; H.U. Hoppe ; R. Luckin ; M. Mavrikis ; K. Porayska-Pomsta ; B. McLaren ; B. d. Boulay. 1. ed. Cham : Springer, 2018. pp. 204-208 (Lecture Notes in Computer Science).
    @inproceedings{0f61243094da4b96a65de8742519a9b5,
    title = "Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks",
    abstract = "Chat conversations between more than two participants are often used in Computer Supported Collaborative Learning (CSCL) scenarios because they enhance collaborative knowledge sharing and sustain creativity. However, multi-participant chats are more difficult to follow and analyze due to the complex ways in which different discussion threads and topics can interact. This paper introduces a novel method based on neural networks for detecting implicit links that uses features computed with string kernels and word embeddings. In contrast to previous experiments with an accuracy of 33{\%}, we obtained a considerable increase to 44{\%} for the same dataset. Our method represents an alternative to more complex deep neural networks that cannot be properly used due to overfitting on limited training data.",
    keywords = "Computer Supported Collaborative Learning, Natural Language Processing, String kernels, Neural networks, Implicit links",
    author = "Mihai Masala and Stefan Ruseti and Gabriel Gutu-Robu and Traian Rebedea and Mihai Dascalu and Stefan Trausan-Matu",
    year = "2018",
    month = "6",
    doi = "10.1007/978-3-319-93846-2_37",
    language = "English",
    isbn = "9783319938455",
    series = "Lecture Notes in Computer Science",
    publisher = "Springer",
    pages = "204--208",
    editor = "C.P. Ros{\'e} and R. Mart{\'i}nez-Maldonado and H.U. Hoppe and R. Luckin and M. Mavrikis and K. Porayska-Pomsta and B. McLaren and Boulay, {B. d. }",
    booktitle = "Artificial Intelligence in Education",
    edition = "1",

    }

    Masala, M, Ruseti, S, Gutu-Robu, G, Rebedea, T, Dascalu, M & Trausan-Matu, S 2018, Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks. in CP Rosé, R Martínez-Maldonado, HU Hoppe, R Luckin, M Mavrikis, K Porayska-Pomsta, B McLaren & BD Boulay (eds), Artificial Intelligence in Education: 19th International Conference, AIED 2018, London, UK, June 27–30, 2018, Proceedings, Part II. 1 edn, Lecture Notes in Computer Science, Springer, Cham, pp. 204-208, Artificial Intelligence in Education, London, United Kingdom, 27/06/18. https://doi.org/10.1007/978-3-319-93846-2_37

    Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks. / Masala, Mihai; Ruseti, Stefan; Gutu-Robu, Gabriel ; Rebedea, Traian; Dascalu, Mihai; Trausan-Matu, Stefan.

    Artificial Intelligence in Education: 19th International Conference, AIED 2018, London, UK, June 27–30, 2018, Proceedings, Part II. ed. / C.P. Rosé; R. Martínez-Maldonado; H.U. Hoppe; R. Luckin; M. Mavrikis; K. Porayska-Pomsta; B. McLaren; B. d. Boulay. 1. ed. Cham : Springer, 2018. p. 204-208 (Lecture Notes in Computer Science).

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

    TY - GEN

    T1 - Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks

    AU - Masala, Mihai

    AU - Ruseti, Stefan

    AU - Gutu-Robu, Gabriel

    AU - Rebedea, Traian

    AU - Dascalu, Mihai

    AU - Trausan-Matu, Stefan

    PY - 2018/6

    Y1 - 2018/6

    N2 - Chat conversations between more than two participants are often used in Computer Supported Collaborative Learning (CSCL) scenarios because they enhance collaborative knowledge sharing and sustain creativity. However, multi-participant chats are more difficult to follow and analyze due to the complex ways in which different discussion threads and topics can interact. This paper introduces a novel method based on neural networks for detecting implicit links that uses features computed with string kernels and word embeddings. In contrast to previous experiments with an accuracy of 33%, we obtained a considerable increase to 44% for the same dataset. Our method represents an alternative to more complex deep neural networks that cannot be properly used due to overfitting on limited training data.

    AB - Chat conversations between more than two participants are often used in Computer Supported Collaborative Learning (CSCL) scenarios because they enhance collaborative knowledge sharing and sustain creativity. However, multi-participant chats are more difficult to follow and analyze due to the complex ways in which different discussion threads and topics can interact. This paper introduces a novel method based on neural networks for detecting implicit links that uses features computed with string kernels and word embeddings. In contrast to previous experiments with an accuracy of 33%, we obtained a considerable increase to 44% for the same dataset. Our method represents an alternative to more complex deep neural networks that cannot be properly used due to overfitting on limited training data.

    KW - Computer Supported Collaborative Learning

    KW - Natural Language Processing

    KW - String kernels

    KW - Neural networks

    KW - Implicit links

    U2 - 10.1007/978-3-319-93846-2_37

    DO - 10.1007/978-3-319-93846-2_37

    M3 - Conference article in proceeding

    SN - 9783319938455

    T3 - Lecture Notes in Computer Science

    SP - 204

    EP - 208

    BT - Artificial Intelligence in Education

    A2 - Rosé, C.P.

    A2 - Martínez-Maldonado, R.

    A2 - Hoppe, H.U.

    A2 - Luckin, R.

    A2 - Mavrikis, M.

    A2 - Porayska-Pomsta, K.

    A2 - McLaren, B.

    A2 - Boulay, B. d.

    PB - Springer

    CY - Cham

    ER -

    Masala M, Ruseti S, Gutu-Robu G, Rebedea T, Dascalu M, Trausan-Matu S. Identifying Implicit Links in CSCL Chats Using String Kernels and Neural Networks. In Rosé CP, Martínez-Maldonado R, Hoppe HU, Luckin R, Mavrikis M, Porayska-Pomsta K, McLaren B, Boulay BD, editors, Artificial Intelligence in Education: 19th International Conference, AIED 2018, London, UK, June 27–30, 2018, Proceedings, Part II. 1 ed. Cham: Springer. 2018. p. 204-208. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-319-93846-2_37