arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1903.05072cs.SIcs.CY
收藏

使用社交媒体描述当地对移民的态度

Characterization of Local Attitudes Toward Immigration Using Social Media

Yerka Freire, Eduardo Graells-Garrido

摘要:

移徙是一种世界性现象,可能在人口中产生不同的反应。人们的态度各不相同,从支持多元文化主义和当地人与外国人之间的交流,到对移民的蔑视和仇恨。由于反移民态度往往体现为暴力和歧视行为,因此必须查明这些态度的特征因素。然而,这样做既昂贵又不切实际,因为传统方法需要付出巨大的努力来收集数据。在本文中,我们建议利用Twitter来描述当地对移民的态度,并以智利为例进行研究,近年来移民人口急剧增加。使用半监督主题建模,我们将49,000名用户定位在从赞成到反对移民的范围内。我们从两个方面描述了这两种态度的特点:与每种态度相关的情感和词汇类别,以及讨论网络结构。我们发现,讨论主要是由海地移民驱动的;讨论的趋势和极性存在时间趋势;网络上的行为与态度不同。这些见解可使决策者了解人们对移徙的感受,并对政策沟通和干预措施的设计产生潜在影响,以改善群体间关系。

英文摘要:

Migration is a worldwide phenomenon that may generate different reactions in the population. Attitudes vary from those that support multiculturalism and communion between locals and foreigners, to contempt and hatred toward immigrants. Since anti-immigration attitudes are often materialized in acts of violence and discrimination, it is important to identify factors that characterize these attitudes. However, doing so is expensive and impractical, as traditional methods require enormous efforts to collect data. In this paper, we propose to leverage Twitter to characterize local attitudes toward immigration, with a case study on Chile, where immigrant population has drastically increased in recent years. Using semi-supervised topic modeling, we situated 49K users into a spectrum ranging from in-favor to against immigration. We characterized both sides of the spectrum in two aspects: the emotions and lexical categories relevant for each attitude, and the discussion network structure. We found that the discussion is mostly driven by Haitian immigration; that there are temporal trends in tendency and polarity of discussion; and that assortative behavior on the network differs with respect to attitude. These insights may inform policy makers on how people feel with respect to migration, with potential implications on communication of policy and the design of interventions to improve inter-group relations.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑