Home // International Journal On Advances in Internet Technology, volume 14, numbers 1 and 2, 2021 // View article

Twitter Search Interface for Looking Back at TV Dramas

Authors:
Taketoshi Ushiama
Haruka Nagai

Keywords: Twitter; social viewing; live-tweeting; TV drama; looking back.

Abstract:
In recent years, while TV dramas are being broadcast, many comments and discussions about the dramas are posted on Twitter. These tweets are called ``live tweets,'' and after watching a drama, users can search for live tweets about scenes of interest to them, enjoy the impressions of other viewers, and deepen their thinking from a different perspective. However, in the current Twitter search function, even if the user searches for a keyword of the target scene, the tweets including the keyword are only presented in sequential order of posting. It takes time for users to find the live tweets of the scene they are interested in. This paper proposes an interface that can efficiently look back at dramas by visualizing the similarity distribution of specific keywords by time for live tweets posted during the drama. In this paper, we propose two Word2Vec-based methods and one TF-IDF-based method to calculate the similarity between keywords and live tweets posted during segments of the drama for visualization. From the results of the evaluation experiments, we found that TF-IDF-based method is the most suitable method for calculating the similarity between keywords and situation segments for visualization. In addition, the results of a usability survey of subjects using the prototype system showed that the proposed interface was able to capture the characteristics of TV drama scenes and was an effective way to look back at TV dramas.

Pages: 73 to 79

Copyright: Copyright (c) to authors, 2021. Used with permission.

Publication date: December 31, 2021

Published in: journal

ISSN: 1942-2652