Navigating the massive world of reddit: using backbone networks to map user interests in social media

Olson, Randal S.; Neal, Zachary P.

doi:10.7717/peerj-cs.4

Cited by 43 publications

(32 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Users in Twitter typically organize themselves around specific interests, such as a sports team or hobby, which facilitates interactions with other users who share similar preferences. These users classify their tweets using topic-specific ''hashtags'' [24]. Tweets containing hashtags are inherently more valuable in terms of embedding extra information and bridging knowledge [35].…”

Section: Tweet-based Analysismentioning

confidence: 99%

From neuroscience to computer science: a topical approach on Twitter

et al. 2017

View full text Add to dashboard Cite

Twitter is perhaps the most influential microblogging service, with 271 million regular users producing approximately 500 million tweets per day. Previous studies of tweets discussing scientific topics are limited to local surveys or may not be representative geographically. This indicates a need to harvest and analyse tweets with the aim of understanding the level of dissemination of science related topics worldwide. In this study, we use Twitter as case of study and explore the hypothesis of science popularization via the social stream. We present and discuss tweets related to popular science around the world using eleven keywords. We analyze a sample of 306,163 tweets posted by 91,557 users with the aim of identifying tweets and those categories formed around temporally similar topics. We systematically examined the data to track and analyze the online activity around users tweeting about popular science. In addition, we identify locations of high Twitter activity that occur in several places around the world. We also examine which sources (mobile devices, apps, and other social networks) are used to share popular science related links. Furthermore, this study provides insights into the geographic density of popular science tweets worldwide. We show that emergent topics related to popular -017-0002-9 science are important because they could help to explore how science becomes of public interest. The study also offers some important insights for studying the type of scientific content that users are more likely to tweet.

show abstract

Section: Tweet-based Analysismentioning

confidence: 99%

From neuroscience to computer science: a topical approach on Twitter

et al. 2017

View full text Add to dashboard Cite

show abstract

“…This is a real world social network and largest dataset available for researchers to study. While Reddit is among the largest online social networks but it has been relatively understudied [9]. The statistics of reddit dataset shown in Table 1, is a collection of anonymous users which are numbered (unique) to give them an identity, followed by the list of his sub-reddit interests (603, pics, gaming, funny, tattoos, Askreddit, Minecraft).…”

Section: Reddit Datasetmentioning

confidence: 99%

Interest based Recommender System for Social Media

Garg¹

2016

IJCA

View full text Add to dashboard Cite

In great popularity of Social Media, flourishing massive world of online social communities entrust users to formulate themselves over set of the interests. Users love to know about the people they like and the different things they are interested into. Thus, they require most appropriate source of information sharing and alert gathering. Recommender system is a solution for it. Every site does have a recommender system but these recommender systems are mostly based on the subscription choices of user"s friends. When a user searches for a particular interest over these sites, the first few results that come up to the screen are the broader interest groups which become over crowded over the time and the user remains with the choice to join an interest group which is of lesser interest than the appropriate one. Hence, this paper proposes a interest based Recommender System named as "MAAC" Recommender System, which recommends the users to follow the more specific interest feeds, solely based on the current subscribed interests of a user instead of demographic knowledge or Friends knowledge. Also, it recommends the interests which are closely related to the interests already linked with the user"s profile. The proposed recommender system proves to be highly effective in recommending the various interests to the user, thus incorporating the high percent accuracy in terms of recall, precision and F1 measure as compared to other recommender system.

show abstract

“…A map of Reddit using commenter co-occurrences has also been previously created using a much smaller sample of comment data (Olsen and Neal, 2015) by treating the co-occurrence matrix as a weighted graph and extracting the network backbone. Relatedly, there has been interest in developing vector representations of graph structures as shown by techniques like DeepWalk (Perozzi et al, 2014) and node2vec (Grover and Leskovec, 2016), which we could potentially use to create additional vector representations to test below.…”

Section: Related Workmentioning

confidence: 99%

community2vec: Vector representations of online communities encode semantic relationships

Martin¹

2017

Proceedings of the Second Workshop on NLP and Computational Social Science

View full text Add to dashboard Cite

Vector embeddings of words have been shown to encode meaningful semantic relationships that enable solving of complex analogies. This vector embedding concept has been extended successfully to many different domains and in this paper we both create and visualize vector representations of an unstructured collection of online communities based on user participation. Further, we quantitatively and qualitatively show that these representations allow solving of semantically meaningful community analogies and also other more general types of relationships. These results could help improve community recommendation engines and also serve as a tool for sociological studies of community relatedness.

show abstract

Navigating the massive world of reddit: using backbone networks to map user interests in social media

Cited by 43 publications

References 29 publications

From neuroscience to computer science: a topical approach on Twitter

From neuroscience to computer science: a topical approach on Twitter

Interest based Recommender System for Social Media

community2vec: Vector representations of online communities encode semantic relationships

Contact Info

Product

Resources

About