{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/event-detection-in-twitter-a-keyword-volume","title":"Event detection in Twitter: A keyword volume approach","arxiv_id":"1901.00570","date":"2019-01-03","proceeding":null,"authors":["Ahmad Hany Hossny","Lewis Mitchell"],"abstract":"Event detection using social media streams needs a set of informative\nfeatures with strong signals that need minimal preprocessing and are highly\nassociated with events of interest. Identifying these informative features as\nkeywords from Twitter is challenging, as people use informal language to\nexpress their thoughts and feelings. This informality includes acronyms,\nmisspelled words, synonyms, transliteration and ambiguous terms. In this paper,\nwe propose an efficient method to select the keywords frequently used in\nTwitter that are mostly associated with events of interest such as protests.\nThe volume of these keywords is tracked in real time to identify the events of\ninterest in a binary classification scheme. We use keywords within word-pairs\nto capture the context. The proposed method is to binarize vectors of daily\ncounts for each word-pair by applying a spike detection temporal filter, then\nuse the Jaccard metric to measure the similarity of the binary vector for each\nword-pair with the binary vector describing event occurrence. The top n\nword-pairs are used as features to classify any day to be an event or non-event\nday. The selected features are tested using multiple classifiers such as Naive\nBayes, SVM, Logistic Regression, KNN and decision trees. They all produced AUC\nROC scores up to 0.91 and F1 scores up to 0.79. The experiment is performed\nusing the English language in multiple cities such as Melbourne, Sydney and\nBrisbane as well as the Indonesian language in Jakarta. The two experiments,\ncomprising different languages and locations, yielded similar results.","url_abs":"http://arxiv.org/abs/1901.00570v1","url_pdf":"http://arxiv.org/pdf/1901.00570v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"event-detection-in-twitter-a-keyword-volume","repo_url":"https://github.com/vsatyav007/repo-eventdetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"binary-classification","task_name":"Binary Classification"},{"task_slug":"event-detection","task_name":"Event Detection"},{"task_slug":"transliteration","task_name":"Transliteration"}],"methods":[{"method_slug":"logistic-regression","method_name":"Logistic Regression"},{"method_slug":"svm","method_name":"SVM"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}