Currently, I have an idea to investigate the human aspects of software development. To me, it will be interesting if I can find the relationship between the contribution of developers in an online discussion such as Stack Overflow and their career as software developers. Since I couldn't find the data which have relationships between both factors, I face a dead end. I have no idea what data I need to extract and where I can find the data.
Any suggestions are welcome.
Thank you.
First of all, are you sure that GitHub's licenses and policies allow you to use the data published therein in such a way?
I can think of several ways to use the data. For example;
- Categorize the projects and create statistics to tell how many industrial engineering / life sciences / etc. projects there are, how their number changes in time, etc.
- Correlate the choice of the programming language with the category,
- Correlate the nationality of the contributors with project categories,
- Number of collaborators and time of development until reaching version 1.0 of a given project.
- And so on...
Seems to me you are putting the cart before the horse! First, decide what YOU want to study THEN think on a plan to get the necessary data and how to analyze it!
First of all, define the objective. For example if you want to make a bug prediction by using data mining technique, here is some feature matrix that maybe will help a lot :
1. Number of revision/commit in a given period of time.
2. Number of the emerging of bugs issue in a given period of time (if available).
3. Line of code per class (average).
4. Number of class in this project.
5. Number of method in each class (average).
6. and etc.
In addition, extracting graph representation of the code structure maybe will give you some ideas to produce a new feature.
- Badges
- Science topic
- Similar topics
- Human-Computer Interaction
- Collaboration