Skip to main content



An Analysis of the network structure of Quora

Paper link: http://dl.acm.org/citation.cfm?id=2488506

 

Quora is a question-and-answer website combining attributes of social network applications such as Facebook, Twitter and some other websites that are based on user voting up content. The co-founders of Quora are two former Facebook employees, who aim at establishing a Q&A site that is really good. Hence they combine what they know about social network when working at Facebook with “conventional” Q&A website to create Quora. Consequently, Quora is born with the gene of social network, and it is really an interesting thing to take a closer look at it.

Quora consist of two major components, users and questions. The network of users is similar to the social network of Twitter—user can follow another user or be followed by others, and they can be friends with other users, too. The network of questions is a hierarchical structure, where questions are labeled and categorized. And there is another key feature of questions network—related questions are interconnected with each other. Besides the networks of users and questions, there is still another network connecting these two components—the user-question network. With this network, there is a new form of relationship between users—answerers and askers are connected to each other by questions; different users sharing a same following topic are connected; users who upvote an answer are also connected. Thus the social network of Quora is more complicated than user-to-user relationship—with the help of asking, answering and voting questions, users actively participate in forming user-question-user relationship with others.

In this paper, authors analyze these three networks in detail. For user network, users who contributed more valuable answers are more likely to have followers and friends. For question network, the structure of network is relatively stable and only a small subset of questions are frequently viewed by people. For user-question network, questions under well-followed topics are more likely to have many viewers or answerers than questions under other topics. Those conclusions are all intuitively reasonable. Based on these three networks, Quora improves its content and user experience. The system will generate user-specific contents on the homepage by mining in the user-question network. And according to question network, the system can recommend some related topics or questions to user. Finally, by listing all relevant questions, the system can help users find contents that interest their friends.

The networks of Quora are interesting to study because they combine the social network with entity network. We may have some questions about such a complicated network—does the concept of strong ties and weak ties still apply to it? Do most of the nodes in the network satisfy the Strong Triadic Closure Property? What we know about the network of Quora is still not enough to answer all these questions, and it is really worth exploring.

 

References:

http://blog.designs.codes/quora-social-network/

http://www.sociobits.org/2015/02/quora-review/2233

https://en.wikipedia.org/wiki/Quora

Comments

2 Responses to “ An Analysis of the network structure of Quora ”

Leave a Reply

Blogging Calendar

September 2015
M T W T F S S
 123456
78910111213
14151617181920
21222324252627
282930  

Archives