Proposal of a System for Visualizing Temporal Changes in Impressions from Tweets
Proposal of a System for Visualizing Temporal Changes in Impressions from Tweets
复制标题
一种用于可视化推文印象时间变化的系统的提案
DOI：
10.1108/ijpcc-02-2015-0011
复制
发表时间：
2015
影响因子：
2.6
通讯作者：
and Tomoya Suzuki
中科院分区：
文献类型：
--
作者：
Tadahiko Kumamoto;Hitomi Wada;and Tomoya Suzuki
目的 – 本文的目的是提出一个 Web 应用程序系统，用于基于 Twitter 用户在 Twitter 上发布的推文所收到的印象的时间变化来可视化 Twitter 用户。设计/方法/方法 – 该系统使用 Twitter API 收集指定用户在指定时间段内发布的推文，使用印象挖掘系统根据三个不同的印象对每条推文进行评级，然后生成饼图和折线图，以使用 Google Chart API 可视化先前处理的结果。发现 – 因为以阴郁主题为主题的新闻文章比对于那些以欢快主题为特色的印象挖掘系统，它使用从报纸数据库创建的印象词典，被认为对于分析负面推文更有效。研究局限性/影响——该系统使用 Twitter API 从 Twitter 收集推文。这表明系统无法收集维护私人时间线的用户的推文。根据我们的调查问卷，大约 30% 的 Twitter 用户的时间线是私人的。这是使用该系统的限制之一。 原创性/价值——该系统使人们能够通过可视化从用户通常在 Twitter 上发布的推文中获得的印象来掌握 Twitter 用户的个性。目标印象仅限于由三个双极印象等级表示的印象：“快乐/悲伤”、“高兴/愤怒”和“平静/紧张”。该系统还使人们能够通过可视化找到关键字的推文的印象来掌握关键字使用的上下文。
Purpose– The purpose of this paper is to propose a Web application system for visualizing Twitter users based on temporal changes in the impressions received from the tweets posted by the users on Twitter.Design/methodology/approach– The system collects a specified user’s tweets posted during a specified period using Twitter API, rates each tweet based on three distinct impressions using an impression mining system, and then generates pie and line charts to visualize results of the previous processing using Google Chart API.Findings– Because there are more news articles featuring somber topics than those featuring cheerful topics, the impression mining system, which uses impression lexicons created from a newspaper database, is considered to be more effective for analyzing negative tweets.Research limitations/implications– The system uses Twitter API to collect tweets from Twitter. This suggests that the system cannot collect tweets of the users who maintain private timelines. According to our questionnaire, about 30 per cent of Twitter users’ timelines are private. This is one of the limitations to using the system.Originality/value– The system enables people to grasp the personality of Twitter users by visualizing the impressions received from tweets the users normally post on Twitter. The target impressions are limited to those represented by three bipolar scales of impressions: “Happy/Sad”, “Glad/Angry” and “Peaceful/Strained”. The system also enables people to grasp the context in which keywords are used by visualizing the impressions from tweets in which the keywords were found.