-
1% rule
-
Community detection
-
Topic modeling the community
-
Influence caused by you and on you
-
Journey of a trending repo
-
Lifetime of a repo
-
Friendship Paradox
-
Degree of separation
- Improving Restaurants by Extracting Subtopics from Yelp Reviews
- Why Your Friends Have More Friends Than You Do
- Friendship Paradox Redux: Your Friends Are More Interesting Than You
- The H-index Paradox: Your Coauthors Have a Higher H-index than You Do
- Generalized friendship paradox in complex networks
- https://arxiv.org/pdf/1407.2535v1.pdf
- http://yuyue.github.io/res/paper/crowdsoft2014.pdf
- Upvote Dynamics on the Quora Network
- Predicting what user reviews are about with LDA and gensim
- Friendship Paradox
- Information Transmission in a Social Network: Dissecting the Spread of a Quora Post
- Geological distribution of the first-degree network connections
- Does your friends have more friends than you?
- User interests based on their (starred) repositories
- Information flow on the GitHub Network
- Actions influenced in you by your followings
- Actions influenced by you in your followers
- What happens when a user(famous/ordinary) creates a repository?
- Famous users can be antirez[current]/tj[bit old]/kennethreitz[more old]
- Ordinary users can be rockstar[contribution faker] owner
- We will compare two repos having ~2000 stars
- People misusing the 'add collaborator' feature to gain attention
- X added Linus Torvald to Y
- Once your repository is at a saturation stage, only bots will star it
- Lifetime of a GitHub repository
- Six degree of separation : small world
- Six degree of separation on (Wikipedia) links on the Internet
- Your collaborators have more commits than you?
- The growth of GitHub, when you joined and what is your id?
- API misuse
- To get more visitors and eventually stargazes on repos.
- They let you add someone as collaborator, and then all of his/her followers get it in their feed, "X added Y to Z".
- How much did they gain by this?
- Role of external channels to make a repository trending or gain eyes
- Hacker News help you more than internal GitHub paths like, search or general random network walk.
- I assume that there is a larger probability of you coming via HN, if you visit in a particular time span after a front page story on HN and if you are not in my level 1 or 2 network.
- **Validate this on a small dataset **
- Cluster people based on their interests
- It deals with the causality, you see a repo and star it, then people from your network do the same.