A method of predicting an evaluation result of an open source community contribution request
A contribution and community technology, applied in the field of predicting the review results of open source community contribution requests, can solve problems such as inaccurate review results, achieve the effect of ensuring accuracy and effectiveness, and speeding up the review cycle
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0040] A specific embodiment of the present invention discloses a method for predicting the review results of open source community contribution requests, such as figure 1 shown, including the following steps:
[0041] Step S1, extract the features of the contribution request to be trained, and generate a feature vector;
[0042] Step S2, using the XGBoost algorithm to construct the prediction model XGPredict;
[0043] Step S3, using the above-mentioned model to predict the review result of the contribution request to be predicted.
[0044] Compared with the prior art, this embodiment provides a method for predicting the review result of an open source community contribution request. After the contributor submits the contribution request, the review result can be quickly estimated. On the one hand, it can enable the contributor to make changes to the code in a timely manner, and on the other hand, it can assist decision makers in making decisions and speed up the review cycl...
Embodiment 2
[0070] Comparing the XGPredict prediction method in Example 1 with the existing RFPredict prediction method, the method in Example 1 removes the attribute features related to comments, and adds several attribute features, which overcomes the existing method RFPredict using contribution requests to submit Then, the shortcomings of attribute characteristics are generated during the review process, and the model is constructed by using these attribute characteristics, so that this method can quickly predict the review result after the contributor submits the contribution request.
[0071] At the same time, the addition of several attribute features makes the prediction accuracy and AUC of the method in Example 1 significantly improved compared with the existing method RFPredict. In the Data Mining challenge of MSR 2014, 90 projects hosted on GitHub were provided. In this example, 28 projects with the largest number of contribution requests were selected from these 90 projects as d...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


