Each metric defined below has its own specific API query. The process is simple : we fetch the data from GitHub APIs which is already streamlined for a specific metric, cut down the data further (if needed) and then assign appropriate weights to these numbers to form a score of a User or a Repo.
A total of 5 metrics have been designed to rate the user and the algorithms are designed according to the process mentioned above.
Each score is a boundless score calculated by summation of weighted sums of each of the following metric.
- Contribution Count
Motive : To get a base idea of a given user
| Weight | Formula |
|---|---|
| 3 | Number of Commits |
| 3 | Number of PRs |
| 2 | Number of Issues |
| 2 | Number of User Repos |
| 3 | Number of Repos contributed to |
- Stars - Repo Ratio
Motive : To get the value of projects created
| Weight | Formula |
|---|---|
| 1 |
|
- PR Acceptance Ratio:
Motive : Value of Contribution
| Weight | Formula |
|---|---|
| 3 |
- Issue Acceptance Ratio:
Motive : Value of Contribution
| Weight | Formula |
|---|---|
| 2 |
- Number of followers/following/organisations involved in. Motive : to get impact in community
| Weight | Formula |
|---|---|
| 10 | Number of Organisations |
| 2 | Number of Followers |
| 1 | Number of Following |
- Annual merged PRs Motive : to get to know how active the user is
| Weight | Formula |
|---|---|
| 3 | Number of PRs merged in an year |
- Annual raised Issues Motive : to get to know how active the user is
| Weight | Formula |
|---|---|
| 2 | Number of PRs merged in an year |
First, skill score is calculated for each topic the User is interested in :
| Weight | Contribution |
|---|---|
| 1 | Number of PRs |
| 1 | Number of Issues |
SKill score for a topic is finally calculated as :