基于机器学习的GitHub企业影响力分析与预测

doi:10.11959/j.issn.2096-6652.202327

智能科学与技术学报 ›› 2023, Vol. 5 ›› Issue (3): 330-342.doi: 10.11959/j.issn.2096-6652.202327

• 专栏：智能科技与社会计算 • 上一篇下一篇

基于机器学习的GitHub企业影响力分析与预测

王明宇¹, 宫庆媛², 瞿晶晶³, 王新¹

¹ 复旦大学计算机科学技术学院，上海 200438
² 复旦大学智能复杂体系基础理论与关键技术实验室，上海 200438
³ 上海人工智能实验室，上海 201210

修回日期:2023-08-02 出版日期:2023-09-01 发布日期:2023-09-26
作者简介:王明宇（1998-），女，2022年在格拉斯哥大学计算机科学学院获得硕士学位，2022—2023年在复旦大学计算机科学技术学院从事科研助理工作。主要研究方向为社交网络、机器学习和大数据分析
宫庆媛（1991- ），女，博士，2020年在复旦大学计算机科学技术学院获得博士学位， 2020—2022年在复旦大学计算机学院继续从事博士后研究工作。2022年起在复旦大学智能复杂体系基础理论与关键技术实验室担任青年副研究员。主要研究方向为在线社交网络用户行为大数据
瞿晶晶（1990- ），女，博士，上海人工智能实验室治理研究中心副研究员。主要研究方向为人工智能创新与治理、科技与社会、科学学
王新（1973- ），男，博士，复旦大学计算机科学技术学院党委书记、教授、博士生导师，主要研究方向为新一代互联网体系结构、无线与移动网络、数据中心网络、社交网络、网络编码的应用
基金资助:
国家自然科学基金项目(62102094）)

Analysis and prediction of GitHub company influence based on machine learning

Mingyu WANG¹, Qingyuan GONG², Jingjing QU³, Xin WANG¹

¹ School of Computer Science, Fudan University, Shanghai 200438, China
² Research Institute of Intelligent Complex Systems, Fudan University, Shanghai 200438, China
³ Shanghai Artificial Intelligent Laboratory, Shanghai 201210, China

Revised:2023-08-02 Online:2023-09-01 Published:2023-09-26
Supported by:
The National Natural Science Foundation of China(62102094）)

摘要/Abstract

摘要：

企业影响力的高低不仅关系到其行业竞争力，也影响着其社会声誉和未来发展，然而对企业影响力的评价一直没有统一的标准。GitHub是一个代表性的软件开发代码存储库开源平台，现有研究通常使用企业在GitHub发布的项目得到的star总数衡量其影响力高低，但是这种方式难以衡量小微企业和新生企业的潜力。通过引入科学家的影响力衡量指标h指数，以GitHub为信息源进行企业网络建模，同时基于该网络提取特征构建分类器，对企业未来的影响力水平进行预测。在此基础上应用SHAP模型解释技术，判别决定企业影响力的重要特征。实验结果显示，基于XGBoost的模型在GitHub真实数据集上实现了0.92的准确率和0.93的平均AUC，可以准确、可靠地对企业进行影响力预测。

关键词: 在线开发者社区, 社交网络, 机器学习, SHAP

Abstract:

The influence of a company is not only related to its industry competitiveness, but also affects its public reputation and future development.However, there has been no unified standard for evaluating the influence of a company.GitHub is a representative open-source platform for software development code repositories.Existing research typically used the total number of stars a company receives for projects posted on GitHub to measure its influence, but this approach is difficult to measure the potential of small, micro, and nascent companies.The paper predicted the future influence level of a company by introducing the scientist's influence measure h-index, using GitHub as the information source, and modeling the company network.Features was extracted features based on this network to build the classifier, which predicted the future influence level of the company.The SHAP model explanation technique was further applied on this basis to identify the important features that determined the influence of a company.The experimental results showed that the XGBoost model achieved an accuracy of 0.92 and an average AUC of 0.93 on the real-world GitHub dataset.In summary, the proposed method could accurately and reliably predict the influence of companies.

Key words: online developer community, social network, machine learning, SHAP

中图分类号:

TP39

王明宇, 宫庆媛, 瞿晶晶, 等. 基于机器学习的GitHub企业影响力分析与预测[J]. 智能科学与技术学报, 2023, 5(3): 330-342.

Mingyu WANG, Qingyuan GONG, Jingjing QU, et al. Analysis and prediction of GitHub company influence based on machine learning[J]. Chinese Journal of Intelligent Science and Technology, 2023, 5(3): 330-342.

图/表 11

表1

图1

图2

表2

图3

图4

表3

图

图6

图7

表4

参考文献 22

[1]	GONG Q Y , LIU Y S , ZHANG J Y ,et al. Detecting malicious accounts in online developer communities using deep learning[C]// Proceedings of IEEE Transactions on Knowledge and Data Engineering. Piscataway:IEEE Press, 2023: 1-17.
[2]	LIU C , YANG D , ZHANG X H ,et al. Recommending GitHub projects for developer onboarding[J]. IEEE Access, 2018,6: 52082-52094.
[3]	李占山, 刘兆赓 . 基于XGBoost的特征选择算法[J]. 通信学报, 2019,40(10): 101-108.
	LI Z S , LIU Z G . Feature selection algorithm based on XGBoost[J]. Journal on Communications, 2019,40(10): 101-108.
[4]	李变 . 基于GitHub社交网络中用户影响力评估算法的研究[D]. 西安:西安电子科技大学, 2015.
	LI B . Research on user influence evaluation algorithm in GitHub-based social networks[D]. Xi’an:Xidian University, 2015.
[5]	HU Y , WANG S , REN Y ,et al. User influence analysis for GitHub developer social networks[J]. Expert Systems with Applications, 2018,108: 108-118.
[6]	BADASHIAN A S , STROULIA E . Measuring user influence in GitHub:the million follower fallacy[C]// Proceedings of the 3rd International Workshop on Crowd Sourcing in Software Engineering. New York:ACM, 2016: 15-21.
[7]	LI S Y , YANG J , LUENG G ,et al. SybilFlyover:heterogeneous graphbased fake account detection model on social networks[J]. KnowledgeBased Systems, 2022,258:110038.
[8]	ZHANG P C , XIONG F , LEUNG H ,et al. FunkR-pDAE:personalized project recommendation using deep learning[J]. IEEE Transactions on Emerging Topics in Computing, 2021,9(2): 886-900.
[9]	BORGES H , HORA A , VALENTE M T . Predicting the popularity of GitHub repositories[C]// Proceedings of the 12th International Conference on Predictive Models and Data Analytics in Software Engineering. New York:ACM, 2016: 1-10.
[10]	VARUNA T V , MOHAN A . Trend prediction of GitHub using time series analysis[C]// Proceedings of 2019 10th International Conference on Computing,Communication and Networking Technologies (ICCCNT). Piscataway:IEEE Press, 2019: 1-7.
[11]	HAN J X , DENG S G , XIA X ,et al. Characterization and prediction of popular projects on GitHub[C]// Proceedings of 2019 IEEE 43rd Annual Computer Software and Applications Conference (COMPSAC). Piscataway:IEEE Press, 2019: 21-26.
[12]	REN L M , SHAN S M , XU X J ,et al. StarIn:an approach to predict the popularity of GitHub repository[M]// Communications in Computer and Information Science. Singapore: Springer Singapore, 2020: 258-273.
[13]	RAHMAN M M , ROY C K , COLLINS J A . CORRECT:code reviewer recommendation in GitHub based on cross-project and technology experience[C]// Proceedings of 2016 IEEE/ACM 38th International Conference on Software Engineering Companion (ICSE-C). Piscataway:IEEE Press, 2017: 222-231.
[14]	LUNDBERG S M , ERION G , CHEN H ,et al. From local explanations to global understanding with explainable AI for trees[J]. Nature Machine Intelligence, 2020,2(1): 56-67.
[15]	贾俊, 张斌, 李志远 ,等. 基于用户行为分析的个性化推荐算法[J]. 智能科学与技术学报, 2019,1(4): 421-426.
	JIA J , ZHANG B , LI Z Y ,et al. Personalized recommendation algorithm based on user behavior analysis[J]. Chinese Journal of Intelligent Science and Technology, 2019,1(4): 421-426.
[16]	THUNG F , BISSYANDé T F , LO D , et al . Network structure of social coding in GitHub[C]// Proceedings of 2013 17th European Conference on Software Maintenance and Reengineering. Piscataway:IEEE Press, 2013: 323-326.
[17]	陈妍, 罗雪琴, 梁伟 ,等. 基于情感信息融合注意力机制的抑郁症识别[J]. 智能科学与技术学报, 2022,4(4): 600-609.
	CHEN Y , LUO X Q , LUENG W ,et al. Depression recognition based on emotional information fused with attentional mechanism[J]. Chinese Journal of Intelligent Science and Technology, 2022,4(4): 600-609.
[18]	陈德旺, 蔡际杰, 黄允浒 . 面向可解释性人工智能与大数据的模糊系统发展展望[J]. 智能科学与技术学报, 2019,1(4): 327-334.
	CHEN D W , CAI J J , HUANG Y H . Development prospect of fuzzy system oriented to interpretable artificial intelligence and big data[J]. Chinese Journal of Intelligent Science and Technology, 2019,1(4): 327-334.
[19]	ZHAO G , DA C , ZOU Y ,et al. Improving the pull requests review process using learning-to-rank algorithms[J]. Empirical Software Engineering, 2019,24(4): 2140-2170.
[20]	RIQUELME F , VERA J A . A parameterizable influence spread-based centrality measure for influential users detection in social networks[J]. Knowledge-Based Systems, 2022,257:109922.
[21]	PANCHENDRARAJAN R , SAXENA A . Topic-based influential user detection:a survey[J]. Applied Intelligence, 2023,53(5): 5998-6024.
[22]	张钹 . 人工智能进入后深度学习时代[J]. 智能科学与技术学报, 2019,1(1): 4-6.
	ZHANG B . Artificial intelligence is entering the post deep-learning era[J]. Chinese Journal of Intelligent Science and Technology, 2019,1(1): 4-6.

企业名称	员工数量/名	创立时间	star总数/次	h指数
Tencent	110 715	1998年	448 913	135
Square	1 853	2009年	281 166	105
IBM	282 100	1911年	66 368	103
Apple	154 000	1976年	217 340	101
Facebook	83 553	2004年	935 897	97
JetBrains	1 900	2000年	161 409	97
Uber	29 000	2009年	98 124	82
Twitter	7 500	2006年	123 067	69
Oracle	143 000	1977年	56 345	64

排名	度量中心性	介数中心性	接近中心性	特征向量中心性
1	'microsoft': 0.522	'microsoft': 0.110	'microsoft': 0.554	'microsoft': 0.164
2	'shopify': 0.470	'shopify': 0.070	'shopify': 0.521	'shopify': 0.162
3	'unity-technologies': 0.406	'unity-technologies': 0.030	'unity-technologies': 0.486	'unity-technologies': 0.158
4	'atlassian': 0.359	'automattic': 0.026	'atlassian': 0.461	'atlassian': 0.151
5	'nokia': 0.346	'atlassian': 0.021	'nokia': 0.456	'nokia': 0.149
6	'github': 0.338	'github': 0.019	'github': 0.454	'github': 0.1 464
7	'apache': 0.326	'apache': 0.019	'apache': 0.446	'apache': 0.144
8	'automattic': 0.282	'nokia': 0.017	'automattic': 0.431	'grafana': 0.139
9	'grafana': 0.280	'mozilla': 0.014	'grafana': 0.429	'segmentio': 0.135
10	'mozilla': 0.274	'JetBrains': 0.009	'mozilla': 0.428	'hashicorp': 0.134
11	'elastic': 0.266	'facebook': 0.009	'elastic': 0.423	'elastic': 0.133
12	'segmentio': 0.264	'intel': 0.008	'segmentio': 0.422	'hubspot': 0.132
13	'facebook': 0.260	'adobe': 0.006	'facebook': 0.421	'appian': 0.131
14	'hashicorp': 0.260	'alibaba': 0.006	'hashicorp': 0.421	'nextdoor': 0.131
15	'hubspot': 0.256	'meteor': 0.006	'hubspot': 0.419	'digitalocean':0.130

预测模型	参数	accuracy	recall	F1-score	precision
XGBoost	learning_rate=0.5, n_estimators=80, min_child_weight=5, subsample=0.8	0.918	0.914	0.889	0.865
SVM	kernel=`linear`, C=16	0.800	0.886	0.861	0.838
MLP	solver=`lbfgs`, hidden_layer_sizes=(5,2), max_iter=1000	0.837	0.714	0.756	0.806

模型	accuracy	recall	F1-score	precision
本文预测模型	0.918	0.914	0.889	0.865
Han等人提出的模型^[11]	0.865	0.816	0.827	0.838

基于机器学习的GitHub企业影响力分析与预测

Analysis and prediction of GitHub company influence based on machine learning

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 22

相关文章 5

Metrics

推荐阅读 0

[1]	许辉. 基于深度Q学习的蜂窝车联网边路资源分配算法[J]. 智能科学与技术学报, 2023, 5(1): 83-91.
[2]	胡东伟, 冯晓璐. 大脑建模的理论框架及热点问题[J]. 智能科学与技术学报, 2021, 3(4): 412-434.
[3]	胡毅, 瞿博阳, 梁静, 王杰, 王艳丽. 进化集成学习算法综述[J]. 智能科学与技术学报, 2021, 3(1): 18-35.
[4]	刘晨熙, 刘炳宏, 张贤, 龙新南, 彭木根. 面向智能服务的雾无线接入网络：原理、技术与挑战[J]. 智能科学与技术学报, 2021, 3(1): 10-17.
[5]	曹汉平,张晓晶,祝睿杰,黄潇拉. 数字金融时代机器学习模型在实时反欺诈中的应用与实践[J]. 智能科学与技术学报, 2019, 1(4): 342-351.