Affiliations: [a] The University of Tokyo, 7-3-1, Hongo, Bunkyoku, Tokyo, Japan. E-mails: [email protected], [email protected] | [b] Hottolink.Inc, 1-3-11 Fujimicho, Chiyodaku, Tokyo, Japan. E-mail: [email protected]
Abstract: Burst phenomena are caused by such social events as flaming on the internet, elections, and natural disasters. To understand people’s thoughts and feelings, we must classify their opinions from burst phenomena. Therefore, classification methods that categorize tweets are critical. However, since most classification methods focus on text mining, they cannot classify tweets by topics because each tweet has poor linguistic similarities. We used a non-text-based method proposed by Baba et al. that groups tweets by topics, even if they have poor linguistic similarities, and verified its validity by comparing it with a text-based method in two different evaluations: full data and sampled data. In the full data evaluation part, we did a questionnaire survey and validated the suitability of the topic clusters created by both classification methods using our full dataset. In the sampled data evaluation part, we focused on the robustness of each method against data reduction. Since collecting the whole data of burst phenomena is very costly due to the vast amounts of available social media data, robustness against data reduction is an important index to evaluate classification methods. After these evaluations, we found that the non-text-based method more effectively classified tweets than the text-based method.
