Analyzing User Behaviors Based on Temporal Patterns of Sequential Pattern Evaluation Indices on Twitter
Analyzing User Behaviors Based on Temporal Patterns of Sequential Pattern Evaluation Indices on Twitter
复制标题
基于 Twitter 上序列模式评估指数的时间模式分析用户行为
DOI：
10.1007/978-3-319-25660-3_15
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Hidenao Abe
中科院分区：
文献类型：
--
作者：
Yusuke Shiozawa;Luca Malcovati;Anna Galli;Aiko Sato-Otsubo;Keisuke Kataoka;Yusuke Sato;Hiromichi Suzuki;Tetsuichi Yoshizato;Kenichi Yoshida;Masashi Sanada;Hideki Makishima;Yuichi Shiraishi;Kenichi Chiba;Eva Hellström Lindberg;Satoru Miy;Hidenao Abe
随着社交媒体网站（诸如Twitter）以推文和推文行为的形式提供用户的日常兴趣和关注的视觉记录，在诸如公司的用户中存在识别其他感兴趣的用户的增长的需求。然而，准确地确定从企业用户接收信息（例如推文）的用户是否对其真正感兴趣可能是困难的。在这项研究中，用户的重发收到的信息在Twitter上（转推）的行为进行了分析，目的是开发一种方法来构建一个模型，预测使用过去的推文历史的内容，通过评价指标的单词和短语的用户的推文的转推行为。本文分析了大型在线零售网站和关注者发送的推文，并将从转发推文中获得的特征词与关注者发送的推文中的特征词进行了比较。本文还讨论了通过提取评价指标的时间模式来构建行为预测模型的可行性，该评价指标是从关注者的推文中获得的特征词和短语的使用频率中创建的。
With social media sites, such as Twitter, providing a visual record of the daily interests and concerns of users in the form of tweets and tweeting behaviors, there is growing demand among users, such as corporations, to identify other interested users. However, accurately determining whether users who receive information (such as tweets) from enterprise users have a genuine interest in it can be difficult. In this study, the user behavior of resending information received on Twitter (retweeting) is analyzed with the aim of developing a method for constructing a model for predicting retweeting behavior using the content of past tweeting history via evaluation indices of words and phrases in the users’ tweets. This paper analyzes the tweets sent by large online retail websites and by the followers who receive them, comparing the feature words obtained from the retweets with those in the tweets sent by the followers. This paper also discusses the feasibility of constructing a behavior prediction model by extracting temporal patterns of evaluation indices that are created from the usage frequencies of feature words and phrases obtained from followers’ tweets.