Coalescing twitter trends: The under-utilization of machine learning in social media

Michael Brennan, Rachel Greenstadt

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    We demonstrate the effectiveness that machine learning can bring to improving social media platforms through a case study on Twitter trending topics. Social media relies heavily on tagging and often does not take advantage of machine learning advances. Twitter is no exception. Individual tweets are identified as being part of a trending discussion topic by the presence of a tagged keyword. Relying solely on this keyword, however, may be inadequate for identifying all the discussion associated with a trend. Our research demonstrates that machine learning techniques can be used identify the top trend a tweet belongs to with up to 85% precision without using the identifying keyword as a feature. This can aid in improving the quality of topic categorization by ensuring on-topic tweets that are missing the trend keyword are included, as well as suggest keywords to include in new tweets.

    Original languageEnglish (US)
    Title of host publicationProceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011
    Pages641-646
    Number of pages6
    DOIs
    StatePublished - Dec 1 2011
    Event2011 IEEE International Conference on Privacy, Security, Risk and Trust, PASSAT 2011 and 2011 IEEE International Conference on Social Computing, SocialCom 2011 - Boston, MA, United States
    Duration: Oct 9 2011Oct 11 2011

    Publication series

    NameProceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011

    Conference

    Conference2011 IEEE International Conference on Privacy, Security, Risk and Trust, PASSAT 2011 and 2011 IEEE International Conference on Social Computing, SocialCom 2011
    CountryUnited States
    CityBoston, MA
    Period10/9/1110/11/11

    Fingerprint

    Learning systems

    ASJC Scopus subject areas

    • Hardware and Architecture
    • Safety, Risk, Reliability and Quality

    Cite this

    Brennan, M., & Greenstadt, R. (2011). Coalescing twitter trends: The under-utilization of machine learning in social media. In Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011 (pp. 641-646). [6113187] (Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011). https://doi.org/10.1109/PASSAT/SocialCom.2011.160

    Coalescing twitter trends : The under-utilization of machine learning in social media. / Brennan, Michael; Greenstadt, Rachel.

    Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011. 2011. p. 641-646 6113187 (Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Brennan, M & Greenstadt, R 2011, Coalescing twitter trends: The under-utilization of machine learning in social media. in Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011., 6113187, Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011, pp. 641-646, 2011 IEEE International Conference on Privacy, Security, Risk and Trust, PASSAT 2011 and 2011 IEEE International Conference on Social Computing, SocialCom 2011, Boston, MA, United States, 10/9/11. https://doi.org/10.1109/PASSAT/SocialCom.2011.160
    Brennan M, Greenstadt R. Coalescing twitter trends: The under-utilization of machine learning in social media. In Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011. 2011. p. 641-646. 6113187. (Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011). https://doi.org/10.1109/PASSAT/SocialCom.2011.160
    Brennan, Michael ; Greenstadt, Rachel. / Coalescing twitter trends : The under-utilization of machine learning in social media. Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011. 2011. pp. 641-646 (Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011).
    @inproceedings{cbac3cd0e60d4e67b2f19ea994a107f1,
    title = "Coalescing twitter trends: The under-utilization of machine learning in social media",
    abstract = "We demonstrate the effectiveness that machine learning can bring to improving social media platforms through a case study on Twitter trending topics. Social media relies heavily on tagging and often does not take advantage of machine learning advances. Twitter is no exception. Individual tweets are identified as being part of a trending discussion topic by the presence of a tagged keyword. Relying solely on this keyword, however, may be inadequate for identifying all the discussion associated with a trend. Our research demonstrates that machine learning techniques can be used identify the top trend a tweet belongs to with up to 85{\%} precision without using the identifying keyword as a feature. This can aid in improving the quality of topic categorization by ensuring on-topic tweets that are missing the trend keyword are included, as well as suggest keywords to include in new tweets.",
    author = "Michael Brennan and Rachel Greenstadt",
    year = "2011",
    month = "12",
    day = "1",
    doi = "10.1109/PASSAT/SocialCom.2011.160",
    language = "English (US)",
    isbn = "9780769545783",
    series = "Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011",
    pages = "641--646",
    booktitle = "Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011",

    }

    TY - GEN

    T1 - Coalescing twitter trends

    T2 - The under-utilization of machine learning in social media

    AU - Brennan, Michael

    AU - Greenstadt, Rachel

    PY - 2011/12/1

    Y1 - 2011/12/1

    N2 - We demonstrate the effectiveness that machine learning can bring to improving social media platforms through a case study on Twitter trending topics. Social media relies heavily on tagging and often does not take advantage of machine learning advances. Twitter is no exception. Individual tweets are identified as being part of a trending discussion topic by the presence of a tagged keyword. Relying solely on this keyword, however, may be inadequate for identifying all the discussion associated with a trend. Our research demonstrates that machine learning techniques can be used identify the top trend a tweet belongs to with up to 85% precision without using the identifying keyword as a feature. This can aid in improving the quality of topic categorization by ensuring on-topic tweets that are missing the trend keyword are included, as well as suggest keywords to include in new tweets.

    AB - We demonstrate the effectiveness that machine learning can bring to improving social media platforms through a case study on Twitter trending topics. Social media relies heavily on tagging and often does not take advantage of machine learning advances. Twitter is no exception. Individual tweets are identified as being part of a trending discussion topic by the presence of a tagged keyword. Relying solely on this keyword, however, may be inadequate for identifying all the discussion associated with a trend. Our research demonstrates that machine learning techniques can be used identify the top trend a tweet belongs to with up to 85% precision without using the identifying keyword as a feature. This can aid in improving the quality of topic categorization by ensuring on-topic tweets that are missing the trend keyword are included, as well as suggest keywords to include in new tweets.

    UR - http://www.scopus.com/inward/record.url?scp=84856135048&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84856135048&partnerID=8YFLogxK

    U2 - 10.1109/PASSAT/SocialCom.2011.160

    DO - 10.1109/PASSAT/SocialCom.2011.160

    M3 - Conference contribution

    AN - SCOPUS:84856135048

    SN - 9780769545783

    T3 - Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011

    SP - 641

    EP - 646

    BT - Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011

    ER -