GitHubNet: Understanding the Characteristics of GitHub Network
Abdullah Talha Kabakus
Department of Computer Engineering, Faculty of Engineering, Duzce University, Turkey
E-mail: talhakabakus@duzce.edu.tr
Received 12 May 2020; Accepted 22 June 2020; Publication 15 September 2020
Abstract
Web 2.0 technologies have not only raised microblogs, but also social software development and collaboration platforms. GitHub is the most popular software development platform that provides social collaboration. Within the scope of this study, a novel graph-based analysis model is proposed which targets to reveal (1) the characteristics of the GitHub in order to shed light on social software development in general, and (2) the most popular programming languages, repositories, and developers in order to shed light on the trending software development technologies. To this end, a subset of the GitHub network, which contains 84, 737 developers and 209, 100 repositories, was collected through the GitHub API and stored on a graph database namely neo4j to be later analyzed. The result of the analysis shows that (1) the connections in GitHub are not mutually linked, (2) JavaScript, Python, and Java are currently the most popular three programming languages, (3) You-Dont-Know-JS, oh-my-zsh, and public-apis are the most popular three repositories, and (4) TarrySingh (Tarry Singh), indrajithban-dara (Indrajith Bandara), and rootsongjc (Jimmy Song) are the most popular three developers. Furthermore, the proposed novel analysis model can be easily applied to other social networks.
Keywords: GitHub, social coding, social network analysis, link analysis, graph database.
1 Introduction
The effects of Web 2.0 were not only emerged in social media, but also in the way the software projects are/were developed. Web 2.0, a.k.a. participatory web, lets users of web send data to the websites which allow users to participate. As a natural consequence of this investment, social networking platforms such as Facebook, Twitter, etc. have been born, where the contents of these platforms are completely generated by the users. GitHub1 is an online platform that lets users share their software projects, which are also called repositories, thanks to the revision control and source code management system, namely Git,2 that it is built on. Alongside publicly sharing the sources of the repositories, GitHub lets users (who are also called developers) collaborate while developing software projects which allow developers from all around the world to contribute to each other’s repositories even they do not know each other in real life. This collaboration is nowadays a necessity for developing software projects that are built on many different technologies (i.e. front-end, backend, etc.) from various parts of software development disciplines (i.e. designing, coding, testing) [1]. GitHub is commonly called a social coding platform as it provides interaction between developers and repositories such as (1) favoring someone’s repository (similar to the ‘like’ action on some several social networks), which is called stargazer in order to show their appreciation, and bookmarking the repository [2], (2) watching repositories to be notified when some changes are applied to them, (3) creating an own exact copy of an existing repository to be able to make changes on it, which is called forking, (4) creating issues regarding the repository, which may led the issue to be resolved since GitHub notifies the repository owner, and (5) the requests by the forkers when they want to commit the changes on their own copies to the original repository, which are called pull requests. These interactions are presented in Figure 1.
Developers can directly communicate with each other on GitHub by adding comments while making commits, creating issues, or making pull requests. In addition to the interactions between developers and repositories, a developer can follow another one in order to see his/her recent public activities (i.e. his/her new commits, recently created issues, and pull requests) on his/her news feed as well as showing respect or support to the developer being followed [3]. Following someone in GitHub is a one-way relationship, which means the developer being followed may not follow his/her follower. There is no limit for a developer to create repositories [4]. GitHub provides a profile page for each developer as it is presented in Figure 2 that hosts an overview of the developer including (1) his/her location, (2) email, (3) website, (4) a brief biography of the developer, (5) the organizations of the developer, (6) repositories owned by him/her, (7) his/her starred repositories, (8) the project boards of the developer, which provide developers to create customized workflows such as comprehensive roadmaps, and release checklists, (9) his/her followers, and (10) the developers s/he follows, described as followings.
Figure 1 The interactions available on GitHub.
Figure 2 A sample profile page of a GitHub developer.
Similar to developers, GitHub provides a profile page for the repositories as it is presented in Figure 3 that hosts (1) the source code of the repository including its commits, branches, releases and contributors, (2) the created issues of the repository, (3) the pull requests of the repository, (4) the project boards of the repository, (5) the insights of the repository, and (6) a wiki for the repository that provides a place in the repository to lay out the roadmap of the project and show the current status of the repository.
Figure 3 A sample profile page of a GitHub repository.
GitHub allows two types of project owners: (1) personal accounts, and (2) organizational accounts. While the former is intended for individuals, the latter is intended for companies or non-profit organizations. The main difference between the personal and organizational accounts is that the organizational accounts cannot have followers [5] as, according to the GitHub, it is currently not supported by design [6]. A personal account can be easily converted into an organizational account but that would make the account lose all its followers.
GitHub is currently the most popular social coding site that hosts more than 96 million repositories from 31+ million developers [7]. The main motivation of this study is the idea of having a good understanding that these huge populations can indeed help to reveal some useful characteristics from GitHub, as a social coding platform. These characteristics include but not limited to both of the software projects and how GitHub is used by the developers. From this point of view, a database was constructed by crawling the GitHub API [8] in order to analyze GitHub through a subset of the entire GitHub network instead of the whole GitHub network as it would be difficult to perform the analysis on the latter due to its size and limitation on human and computation resources. The main contributions of this study are listed as follows:
- An up-to-date subset of GitHub network, which is publicly available3, was constructed by preserving the connections, which model the interactions available in GitHub.
- The trending programming languages and software technologies were revealed by analyzing the constructed dataset.
- The nature of the connections between the developers in GitHub was revealed through the experiments conducted.
- The most influential developers and repositories were revealed by utilizing centrality measurements on the graph model constructed.
- A novel graph-based analysis model that can be easily adapted to any social networks was proposed.
The rest of the paper is structured as follows: Section 2 briefly describes the related work. Section 3 presents the material and method used by the proposed study. Section 4 presents the experimental results and discussion. Finally, Section 5 concludes the paper with future directions.
2 Related Work
Hu et al. [9] proposed an analysis of the influences of GitHub repositories on the basis of the star connection between developers and repositories. As a result of the analysis, they revealed the most influential repositories on GitHub. Thung et al. [10] collected 100, 000 repositories and 30, 000 developers in order to compute various characteristics of the developer-developer and repository-repository graphs constructed. In the wake of their experiments, they deduced that project networks were more interconnected than the developer networks.
Jiang et al. [11] analyzed the dissemination of the repositories in GitHub in terms of the speed and range on the dataset collected, which consisted of 2, 665 repositories and 747, 107 developers. According to the experimental results, they highlighted that social connections were not reciprocal, and the social links played a notable role in repository dissemination.
Ray et al. [12] carried out a study to shed light on the question of “What is the effect of the programming language on software quality?” through GitHub. To this end, they collected 729 repositories in 17 different languages from GitHub. According to the experimental results, their observations with regard to the effect of the language design are such that: (1) Strong typing is modestly better than weak typing, (2) static typing is better than dynamic typing, and (3) functional programming languages are better than procedural languages. They clearly noted that the effects arising from the design of the language were dominated by the process factors such as the size of the project, the size of the project team, and the commit size. Borges et al. [2] analyzed the insights of the starring practices in GitHub through the 5, 000 repositories collected and conducted surveys on the developers. They concluded their study with notable conclusions as such that stars were viewed as the most useful measure of popularity in GitHub, and developers in GitHub considered the number of stars of the repository before using or contributing to the repository.
Dabbish et al. [3] studied how the developers in GitHub inferred other developers’ technical goals and vision, which eventually advanced their technical skills and knowledge. The methodology they used was conducting a series of semi-structured interviews with 24 GitHub developers, which were documented in order to be analyzed in detail. Rusk and Coady [13] proposed an approach that characterized the developer activity by the geographic location with an aim to identify (1) the regional skilled professionals for employers, and (2) which local skills would be in greatest demand for developers. Jiang et al. [6] investigated the unfollowing behavior on GitHub. To this end, they constructed a dataset that contained 701, 364 developers. According to their experiments, they concluded the study that developers were more likely to unfollow those who had had fewer activities, lower programming language similarity, and asymmetric connections.
Vasilescu et al. [14] analyzed the impact of social diversity within the project teams on the software development process. To this end, they collected a team social diversity data set of 23, 493 repositories from GitHub. They used some measurements such as the genders and the countries of team members, the number of contributors leaving, joining, and staying in the team as well as the turnover ratio in order to analyze the impact of the social diversity. Yu et al. [15] analyzed the growth curves of GitHub compared with the traditional open source communities in order to reveal why GitHub grew in an explosive way. They also studied the follow-network of the developers and presented various patterns of social behavior.
Apart from GitHub, some researchers [16, 17] investigated the use of microblogs (i.e. Twitter4) in software development. Treude et al. [18] investigated the way Stack Overflow5, which is a quite popular question and answer platform for developers, is used by developers when developing software projects. Surian et al. [19] studied the network structure of SourceForge6, another popular online software development and collaboration platform.
3 Material and Method
GitHub provides a REST API7 to query the data stored on its own database. Within the scope of this study, an up-to-date subset of the GitHub network, namely, GitHubNet, was constructed using this REST API of GitHub. Since social networks (i.e. Twitter, Facebook, Instagram, etc.) can be effectively modeled as graphs, the constructed dataset was stored on a graph database, namely neo4j8, where the entities were modeled as the nodes, and the connections between entities were modeled as the edges between nodes. The following graphs were constructed in order to investigate the insights of the GitHub network.
3.1 Developer-Developer Network
Developer-Developer (D-D) network is a directed graph of developers. The direct interaction between developers that GitHub provides is the “follow” action, which lets developers follow other developers. This “follow” connection does not have to be mutual. Therefore, the following action was defined by the FOLLOWS connection from the developer to the developer, who was being followed. Developers who contribute in at least one same project were identified as “collaborated developers” and a connection named COLLABORATES that indicates the collaboration between two developers was established. This connection was established for each developer. The developers, whose descriptions on GitHub contained the “developer” keyword, were retrieved from GitHub API and the D – D network was constructed. The pseudo-code of the algorithm that is used to construct the D – D network is presented in Algorithm 1.
The constructed connections of the D – D network, which are FOLLOWS, and COLLABORATES, are illustrated in Figure 4.
3.2 Developer-Repository Network
Developer-Repository (D-R) network is a bipartite type of network where the nodes represent the developers and the repositories. The direct interactions between a developer and a repository are (1) developers may create repositories, which are named “owned repositories”, (2) developers may mark repositories as favorite, which is also known as stargazer, (3) developers may watch the repositories in order to be informed when they are updated, and (4) developers may fork others’ repositories in order to create the same copy of the repository that is open for making changes on it. Therefore, the D – R network included a subnetwork that contains the FORK connections from the original repository to the forked one. Alongside owning a repository, a developer may contribute to others’ repositories by submitting changes in the code (a.k.a. commits), creating issues regarding the project, making feature requests, and proposing pull requests in order to submit the changes on the owned repository to the original one. The repositories that each developer contributed to were retrieved from the GitHub API and CONTRIBUTES connections were established between the developers and the repositories they contributed to. The pseudo-code of the algorithm that was used to construct the D – R network is presented in Algorithm 2.
Algorithm 1 The pseudo-code of the algorithm that is used to construct the D – D network.
Figure 4 The constructed connections of the D – D network.
The constructed connections of the D – R network, which are OWNS, STARS, WATCHES, CONTRIBUTES, and FORKS, are illustrated in Figure 5.
Algorithm 2 The pseudo-code of the algorithm that is used to construct the D – R network.
Figure 5 The constructed connections of the D – R network.
3.3 Repository-Language Network
Repository-Language (R-L) network is a bipartite type of network where the nodes represent the repositories and the programming languages. GitHub defines a primary language for each repository. Even though a repository contains source code from different programming languages (i.e. dynamic web applications, native mobile applications, etc.), the primary programming language is set as the language of the repository. The pseudo-code of the algorithm that is used to construct the R – L network is presented in Algorithm 3.
Algorithm 3 The pseudo-code of the algorithm that is used to construct the R – L network.
Figure 6 The constructed connection of the R – L network.
The constructed connection of the R – L network, which is LANG, is illustrated in Figure 6.
4 Experimental Results and Discussion
In order to identify the insights of the GitHub network, a large subset of the GitHub network was collected from the GitHub API thanks to the implemented Python scripts. The developers were queried through their descriptions as long as their descriptions contain the keyword “developer”, they were collected. A generic keyword such as “developer” was intentionally selected to construct a comprehensive network, that covers various developers. Based on this set, the repositories of these developers were formed the repository set. Consequently, 84, 737 developers and 205, 229 repositories were collected. Since different connections establish different types of entities, which are available in GitHub, three different graph databases were constructed. Each graph database consisted of its own entities (nodes) and the connections (edges) between these entities. In order to understand the characteristics of each network, several measurements are needed to be calculated as follows: (1) closeness centrality, which defines the sum of the length of the shortest path between the node and all other nodes, (2) harmonic centrality, which is a variant of closeness centrality, and (3) average clustering coefficient, which is the mean of local clustering. In order to weigh the importance of web-pages based on their links as a part of the Google9 search engine, the PageRank algorithm was proposed by Brin and Page [20]. The most influential and central nodes for each network were identified using PageRank, closeness centrality, and harmonic centrality measurements. In addition to that, the average clustering coefficient was also calculated for each network in order to reveal the mean of the local clustering. For the D-D network, the PageRank algorithm reveals the most influential developers since all connections in this network were established with developer nodes. Table 1 lists the analysis result of the D - D network. According to the analysis result, TarrySingh (Tarry Singh),10 indrajithbandara (Indrajith Bandara),11 and rootsongjc (Jimmy Song)12 were the three most popular developers.
Table 1 The analysis result of the D – D network
| # | PageRank | Closeness Centrality | Harmonic Centrality |
| 1 | TarrySingh | gitter-badger | gitter-badger |
| 2 | indrajithbandara | pra85 | TarrySingh |
| 3 | rootsongjc | TarrySingh | indrajithbandara |
| 4 | mitsuhiko | mitsuhiko | pra85 |
| 5 | amueller | benbalter | invalid-email-address |
| 6 | onevcat | invalid-email-address | mitsuhiko |
| 7 | alvarotrigo | indrajithbandara | ReadmeCritic |
| 8 | ryanflorence | amueller | amueller |
| 9 | fengmk2 | toddmotto | pborreli |
| 10 | jamesmontemagno | ReadmeCritic | benbalter |
| Average Clustering Coefficient: 0.04967589042886072 | |||
The PageRank algorithm ranks the nodes (webpages) according to incoming links to them [21]. The higher the indegree of a node is, the higher the rank the node would have. Since all connections were established with repository nodes in the D – R network, the PageRank algorithm revealed the most influential repositories in this network as they are listed in Table 2. In addition to that, most central developers and repositories were revealed. According to the analysis result, You-Dont-Know-JS13, oh-my-zsh14, and public-apis15 were the three most popular repositories. The neighborhood of the D-R network was not complete as the average clustering coefficient was calculated as low as 0.0002476306626778342 as it measures that how complete is a node’s neighborhood, and the clustering coefficient of a network is the average of clustering coefficient of all the nodes in the network [22].
Table 2 The analysis result of the D – R network
| # | PageRank | Closeness Centrality | Harmonic Centrality |
| 1 | getify/You-Dont-Know-JS | xujihui1985 | getify/You-Dont-Know-JS |
| 2 | robbyrussell/oh-my-zsh | getify/You-Dont-Know-JS | robbyrussell/oh-my-zsh |
| 3 | toddmotto/public-apis | toddmotto/public-apis | xujihui1985 |
| 4 | erikras/react-redux-universal-hot-example | trimstray/nginx-quick-reference | toddmotto/public-apis |
| 5 | onevcat/VVDocumenter-Xcode | imthenachoman/How-To-Secure-A-Linux-Server | getify |
| 6 | erikras/redux-form | sveinbjornt/Sloth | jefflau |
| 7 | onevcat/Kingfisher | mitsuhiko/pipsi | nikolay |
| 8 | tangqiaoboy/iOSBlogCN | eranyanay/1m-go-websockets | imthenachoman/How-To-Secure-A-Linux-Server |
| 9 | getify/Functional-Light-JS | leon-ai/leon | trimstray/nginx-quick-reference |
| 10 | kymjs/KJFrameForAndroid | denoland/deno | mitsuhiko/pipsi |
| Average Clustering Coefficient: 0.0002476306626778342 | |||
Since all connections are established through programming language nodes in the R – L network, the PageRank algorithm reveals the most influential programming languages in the network as they are listed in Table 3. In addition to that, the repositories and programming languages were revealed through their centralities. According to the analysis result, JavaScript, Python, and Java were the three most popular programming languages when the GitHub knowledge-base was queried for the keyword “developer”. Similar to the D – R network, the neighborhood of the R – L network was not complete as the average clustering coefficient was calculated as low as 0.
Table 3 The analysis result of the R-L network
| # | PageRank | Closeness Centrality | Harmonic Centrality |
| 1 | JavaScript | JavaScript | JavaScript |
| 2 | Python | getify/A-Tale-Of-Three-Lists | getify/A-Tale-Of-Three-Lists |
| 3 | Java | getify/asyncGate.js | getify/asyncGate.js |
| 4 | PHP | getify/asynquence | getify/asynquence |
| 5 | HTML | getify/breakthewebforward.com | getify/breakthewebforward.com |
| 6 | Ruby | getify/CAF | getify/CAF |
| 7 | CSS | getify/cloud-sweeper | getify/cloud-sweeper |
| 8 | C# | getify/deePool | getify/deePool |
| 9 | Objective-C | getify/DOMEventBridge | getify/DOMEventBridge |
| 10 | TypeScript | getify/ES-Feature-Tests | getify/ES-Feature-Tests |
| Average Clustering Coefficient: 0.0 | |||
As described in the related work in Section 2, there are some studies that provide analysis regarding software development over social networks including GitHub, Stack Overflow, SourceForge, and even Twitter, which does not have a purpose to aid the software development process. A comparison of the related work to the proposed one, namely GitHubNet, in terms of (1) the number of developers each dataset contains, (2) the number of repositories each dataset contains, and (3) the year of each dataset constructed is listed in Table 4.
Each GitHub repository may contain a README file (commonly named as README.md) that describes the project proposed in the repository. This file commonly contains some technical background regarding repositories. The descriptions of the collected repositories were fetched from the GitHub API in order to reveal the trending technologies and topics from the insights. Since these README files contain markup symbols that define the styles of the text, the fetched text was cleared from these symbols as well as punctuation marks, digits and the stop words in English (e.g. “the”, “a”, “an”, “is”, etc.). When the descriptions of the 85, 734 original repositories (the repositories which are not forked from another repository) were analyzed, the notable words, that give clue about the most demanding software development technologies, were found as JavaScript, which is the most popular programming language according to the Stack Overflow Developer Survey 2019 [24], Node which is the most popular web framework based on JavaScript [24], and React, which is the 4th most popular repository16 on GitHub alongside the 96+ million repositories. The word cloud of the analysis is presented in Figure 7.
Table 4 A comparison of the proposed work with the related work
| Related Work | Number of Developers | Number of Repositories | Year of the Dataset Constru |
| Jiang et al. [11] | 747, 107 | 2, 665 | 2013 |
| Luo et al. [1] | 499, 485 | 108, 718 | 2014 |
| Surian et al. [19] | 55, 694 | 28, 087 | 2009 |
| Bana and Arora [22] | 8, 000 | NA | 2018 |
| Thung et al. [10] | 30, 000 | 100, 000 | 2013 |
| Leibzon [23] | NA | NA | 2016 |
| Vasilescu et al. [14] | NA | 23, 493 | 2014 |
| Jiang et al. [6] | 701, 364 | NA | 2016 |
| GitHubNet | 84, 737 | 209,100 | 2019 |
Figure 7 The word cloud of the descriptions of the 85, 734 original repositories.
5 Conclusion
Web 2.0 technologies have not only raised microblogs and social networking sites, but also social software development and collaboration platforms. GitHub is the most popular software development platform that provides social collaboration. Within the scope of this study, a novel graph-based analysis model is proposed which targets to reveal (1) the insights of the GitHub in order to shed light on social software development in general, and (2) the most popular programming languages, repositories, and developers in order to shed light on the trending software development technologies. For this reason, a subset of the GitHub network, which contained 84, 737 developers and 209, 100 repositories, was collected through the GitHub API by querying for the keyword “developer” and stored on a graph database namely neo4j to perform analysis on it. The result of the analysis shows that (1) the connections in GitHub are not mutually linked, (2) JavaScript, Python, and Java are currently the most popular three programming languages, (3) You-Dont-Know-JS, oh-my-zsh, and public-apis are the most popular three repositories, and (4) TarrySingh (Tarry Singh), indrajithbandara (Indrajith Bandara), and rootsongjc (Jimmy Song) are the most popular three developers.
As future work, longitudinal research to investigate how the GitHub network structure evolves over time can be studied. To this end, various snapshots of the GitHub network are needed to be collected and analyzed over time.
1https://github.com
2https://git-scm.com
3https://www.kaggle.com/talhakabakus/githubnet
4https://twitter.com
5https://stackoverflow.com
6https://sourceforge.net
7https://developer.github.com/v3/
8https://neo4j.com
9https://www.google.com
10https://github.com/TarrySingh
11https://github.com/indrajithbandara
12https://github.com/rootsongjc
13https://github.com/getify/You-Dont-Know-JS
14https://github.com/robbyrussell/oh-my-zsh
15https://github.com/toddmotto/public-apis
16According to the http://gitmostwanted.com/top/stars/ and based on the number of stars.
References
[1] Z. Luo, X. Mao, and A. Li, “An Exploratory Research of GitHub Based on Graph Model,” in Proceedings – 2015 9th International Conference on Frontier of Computer Science and Technology, FCST 2015, 2015, pp. 96–103.
[2] H. Borges and M. Tulio Valente, “What’s in a GitHub Star? Understanding Repository Starring Practices in a Social Coding Platform,” J. Syst. Softw., vol. 146, pp. 112–129, 2018.
[3] L. Dabbish, C. Stuart, J. Tsay, and J. Herbsleb, “Social Coding in GitHub: Transparency and Collaboration in an Open Software Repository,” in Proceedings of the ACM 2012 conference on Computer Supported Cooperative Work (CSCW ’12), 2012, pp. 1277–1286.
[4] “Pricing – Plans for every developer,” GitHub, 2019. [Online]. Available: https://github.com/pricing. [Accessed: 06-Dec-2019].
[5] K. Blincoe, J. Sheoran, S. Goggins, E. Petakovic, and D. Damian, “Understanding the popular users: Following, affiliation influence and leadership on GitHub,” Inf. Softw. Technol., vol. 70, pp. 30–39, 2016.
[6] J. Jiang, D. Lo, Y. Yang, J. Li, and L. Zhang, “A first look at unfollowing behavior on GitHub,” Inf. Softw. Technol., vol. 105, pp. 150–160, 2019.
[7] “The State of the Octoverse,” GitHub, 2019. [Online]. Available: https://octoverse.github.com. [Accessed: 06-Dec-2019].
[8] “GitHub API v3 | GitHub Developer Guide,” GitHub, 2019. [Online]. Available: https://developer.github.com/v3/. [Accessed: 06-Dec-2019].
[9] Y. Hu, J. Zhang, X. Bai, S. Yu, and Z. Yang, “Influence analysis of Github repositories,” Springerplus, vol. 5, no. 1, pp. 1–19, 2016.
[10] F. Thung, T. F. Bissyande’, D. Lo, and L. Jiang, “Network structure of social coding in GitHub,” in Proceedings of the 17th European Conference on Software Maintenance and Reengineering (CSMR 2013), 2013, pp. 323–326.
[11] J. Jiang, L. Zhang, and L. Li, “Understanding Project Dissemination on a Social Coding Site,” in Proceedings of 20th Working Conference on Reverse Engineering (WCRE 2013), 2013, pp. 132–141.
[12] B. Ray, D. Posnett, P. Devanbu, and V. Filkov, “A large-scale study of programming languages and code quality in GitHub,” Commun. ACM, vol. 60, no. 10, pp. 91–100, 2017.
[13] D. Rusk and Y. Coady, “Location-based analysis of developers and technologies on GitHub,” in Proceedings of 2014 28th International Conference on Advanced Information Networking and Applications Workshops, 2014, pp. 681–685.
[14] B. Vasilescu, A. Serebrenik, and V. Filkov, “A Data Set for Social Diversity Studies of GitHub Teams,” in Proceedings of 2015 IEEE/ACM 12th Working Conference on Mining Software Repositories, 2015, pp. 514–517.
[15] Y. Yu, G. Yin, H. Wang, and T. Wang, “Exploring the patterns of social behavior in GitHub,” in Proceedings of the 1st International Workshop on Crowd-based Software Development Methods and Technologies (CrowdSoft 2014), 2014, pp. 31–36.
[16] G. Bougie, J. Starke, M.-A. Storey, and D. M. German, “Towards understanding twitter use in software engineering: Preliminary findings, ongoing challenges and future questions,” in Proceedings of the 2nd International Workshop on Web 2.0 for Software Engineering (Web2SE ’11), 2011, pp. 31–36.
[17] Y. Tian, P. Achananuparp, I. N. Lubis, D. Lo, and E. P. Lim, “What does software engineering community microblog about?,” in IEEE International Working Conference on Mining Software Repositories, 2012, pp. 247–250.
[18] C. Treude, O. Barzilay, and M.-A. Storey, “How do programmers ask and answer questions on the web?: NIER track,” in Proceeding of the 33rd international conference on Software engineering (ICSE ’11), 2011, pp. 804–807.
[19] D. Surian, D. Lo, and E. P. Lim, “Mining Collaboration Patterns from a Large Developer Network,” in Proceedings of 2010 17th Working Conference on Reverse Engineering, 2010, pp. 269–273.
[20] S. Brin and L. Page, “The anatomy of a large scale hypertextual Web search engine,” Comput. Networks ISDN Syst., vol. 30, pp. 107–117, 1998.
[21] E. Agirre and A. Soroa, “Personalizing PageRank for word sense disambiguation,” in Proceedings of the 12th Conference of the European Chapter of the ACL (EACL 2009), 2009, pp. 33–41.
[22] R. Bana and A. Arora, “Influence Indexing of Developers, Repositories, Technologies and Programming languages on Social Coding Community GitHub,” in 2018 Eleventh International Conference on Contemporary Computing (IC3), 2018, pp. 1–6.
[23] W. Leibzon, “Social network of software development at GitHub,” in Proceedings of the 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM 2016), 2016, pp. 1374–1376.
[24] “Stack Overflow Developer Survey 2019,” Stack Overflow, 2019. [Online]. Available: https://insights.stackoverflow.com/survey/2019. [Accessed: 06-Dec-2019].
Biography
Abdullah Talha Kabakus received the bachelor’s degree in computer engineering from Cankaya University in 2010, the master’s degree in computer engineering from Gazi University in 2014, and the philosophy of doctorate degree in Electrical-Electronics & Computer Engineering from Duzce University in 2017, respectively. He is currently working as an Assistant Professor at the Department of Computer Engineering, Faculty of Engineering, Duzce University. His research areas include mobile security, deep learning, and social network analysis. He has been serving as a reviewer for many highly-respected journals.
Journal of Web Engineering, Vol. 19_5–6, 557–574.
doi: 10.13052/jwe1540-9589.19561
© 2020 River Publishers
