Open Source Software Supply Chain Recommendation Based on Heterogeneous Information Network

  • Hai Ming Lin
  • , Guanyu Liang
  • , Yanjun Wu
  • , Bin Wu
  • , Chunqi Tian*
  • , Wei Wang
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In the GitHub open-source collaborative development scenario, each entity type and the link relationship between them have natural heterogeneous attributes. In order to improve the accuracy of project recommendation, it is necessary to effectively integrate this multi-source information. Therefore, for the project recommendation scenario, this paper defines an open source weighted heterogeneous information network to represent the different entity types and link relationships in the GitHub open source collaborative development scenario, and effectively model the complex interaction among developers, projects and other entities. Using the weighted heterogeneous information network embedding method, extract and use the rich structural and semantic information in the weighted heterogeneous open source information network to learn the node representation of developers and projects, and fuse the personalized nonlinear fusion function into the matrix decomposition model for open source project recommendation. Finally, this paper makes a large number of comparative experiments based on the real GitHub open data set, and compares it with other project recommendation methods to verify the effectiveness of our proposed open source project recommendation model. At the same time, it also explores the impact of different metapaths on the effect of project recommendation. The experimental results show that the recommendation method based on heterogeneous information network can effectively improve the recommendation quality.

Original languageEnglish
Title of host publicationBenchmarking, Measuring, and Optimizing - 14th Bench Council International Symposium, Bench 2022, Revised Selected Papers
EditorsAna Gainaru, Ce Zhang, Chunjie Luo
PublisherSpringer Science and Business Media Deutschland GmbH
Pages70-86
Number of pages17
ISBN (Print)9783031311796
DOIs
StatePublished - 2023
Event14th International Symposium on Benchmarking, Measuring, and Optimization, Bench 2022 - Virtual, Online
Duration: 7 Nov 20229 Nov 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13852 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference14th International Symposium on Benchmarking, Measuring, and Optimization, Bench 2022
CityVirtual, Online
Period7/11/229/11/22

Keywords

  • GitHub project recommendation
  • Heterogeneous Information Network
  • Open source

Fingerprint

Dive into the research topics of 'Open Source Software Supply Chain Recommendation Based on Heterogeneous Information Network'. Together they form a unique fingerprint.

Cite this