Understanding How People Attend to and Engage with Foreign Language Posts in Multilingual Newsfeeds

We explore how language affects people’s attention to and engagement with social media posts in an eye-tracking experiment in which participants viewed a mock newsfeed containing English and foreign language posts, half with and half without images. Participants spent less time looking at foreign language posts than English posts, and more time looking at posts with images than those without images. Participants reported being more likely to like or comment on posts in English and posts with images. We suggest some new design ideas for supporting people’s interaction on multilingual social media sites. Introduction and Research Background Social media sites provide new opportunities to encounter, understand and interact with linguistically or culturally diverse social connections (Eleta and Golbeck, 2012; Ellison, Wohn, and Greenhow, 2014; Lee and Markey, 2014). For example, Hong, Convertino, and Chi (2011) found that more than 100 languages were used on Twitter. Previous work has focused on the posts by bilingual users, for example looking at how audience and communicative goals influence language choice (Halim and Maros, 2014; Tang et al., 2011). However, less work has considered how people consume and engage with posts in different language in their social media feeds. In this paper, we use eye-tracking to understand how people attend to and engage with foreign language vs. English posts with and without images. People might automatically filter out posts in languages they do not understand, in an early stage of cognitive processing (Treisman, 1969). Alternatively, people might try to process foreign language posts but cut this processing short due to the small amount of information that can extract from such posts. Either way, we expect that people will spend less time on posts in foreign languages than posts in English. Copyright © 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. H1: People will spend more time viewing social media posts in English than posts in foreign languages The amount of information a monolingual English speaker can obtain from a foreign language post, and thus the amount of attention allocated to that post, may also be influenced by structural features of the post such as the inclusion of images. Images have previously been shown to increase attention to posts (Vraga, Bode, and TrollerRenfree, 2016). We expect to replicate these results and then we explore how post language and modality interact. H2: People will spend more time on viewing social media posts including images than posts without images RQ1: How will post language and post modality interact with one’s attention to social media posts? Furthermore, we explore how post language influence on the interestingness of a post. Possibly, native English speakers might find foreign language posts interesting because they are relatively unusual. Also, English speakers might find foreign language posts uninteresting due to their inability to read them. We therefore ask, RQ2: How will post language affect one’s evaluation of interestingness of social media posts? Beyond passive browsing of social media posts, people can contribute to others’ contents such as clicking Like or leaving comments. Previous literature explored what factors increase engagements with posts and found that structural characteristics (Counts and Fisher, 2011; Suh, Hong, Pirolli, and Chi, 2010) predict greater engagement. Because language is also an important structural characteristic of messages, we explore how differently people would engage with posts depending on post language and modality. RQ3: How will post language and modality affect one’s willingness to engage with posts? In addition, we presume that the amount of information and attention allocated to foreign posts may be different depending on the similarity between that language and English. For example, there may be recognizable words in Proceedings of the Eleventh International AAAI Conference on Web and Social Media (ICWSM 2017)