Download PDFOpen PDF in browser

Realization of Discovery for Burst Topic Transition Using the Topic Change Point Detection Method for Time-Series Text Data

11 pagesPublished: September 20, 2022

Abstract

In this paper, we present a realization method of discovery for burst topic transition using the topic change point detection method for time-series text data. In our method, we focus on the topic change point detection method for time-series text data. By similarity measure using the topic change point detection method for time-series text data, we can discover for burst topic transition. In general, when we would like to understand the outline or main points of an event, we often read articles written by people who know information about the event or ask others who are aware of the event to tell us about it. However, the information obtained by these means is hearsay from others and subject to third-party bias, it is difficult to comprehend the events objectively. In our paper, we focus on the topic change and extract the topic change point detection It enables us to discover burst topic transitions. In this paper, we describe an evaluation experiment of a prototype system using our discovery for burst topic transition to verify the effectiveness of our method. We also implement an application by the user interface that provides some crews of a trendy word.

Keyphrases: change point detection, SNS data, time-series text data, topic model, topic transition

In: Tokuro Matsuo (editor). Proceedings of 11th International Congress on Advanced Applied Informatics, vol 81, pages 362--372

Links:
BibTeX entry
@inproceedings{IIAIAAI2021-Winter:Realization_of_Discovery_for,
  author    = {Yuta Ishii and Aiha Ikegami and Takafumi Nakanishi},
  title     = {Realization of Discovery for Burst Topic Transition Using the Topic Change Point Detection Method for Time-Series Text Data},
  booktitle = {Proceedings of 11th International Congress on Advanced Applied Informatics},
  editor    = {Tokuro Matsuo},
  series    = {EPiC Series in Computing},
  volume    = {81},
  pages     = {362--372},
  year      = {2022},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2398-7340},
  url       = {https://easychair.org/publications/paper/ZmZd},
  doi       = {10.29007/k8pb}}
Download PDFOpen PDF in browser