Summary of Lee et al.’s “The Role of Openness in Open Collaboration: A Focus on Open‐Source Software Development Projects”

TO: Professor Ellis

FROM: Teodor Barbu

DATE: September 23, 2020

SUBJECT: 500-Word Summary

This memo is a 500-word summary of the article, “The Role of Openness in Open Collaboration: A Focus on Open‐Source Software Development Projects,” by Saerom Lee, Hyunmi Baek, and Sehwan Oh, professors at the University in Deagu and Seoul, Republic of Korea.

Easy access to information facilitated groups of people working on open projects over the internet and innovative companies learned how to exploit this as an important tool. GitHub and SourceForge are two platforms where people can open projects and developers work together to achieve a common goal. Open-source software development (OSSD) became an alternative to getting knowledge from the outside of the organization in a way to benefit both the organizations and developers. In this article the authors conduct an experiment to determine how exploration and exploitation, concepts of organizational learning, impact the development of a project. To effectively tackle the problems of OSSD, developers were separated in categories that either explore the internal resources or exploit the outside knowledge. For this research data was gathered from 17,691 repositories from GitHub. A team of developers, called an organization, can work on one or more projects, and can collaborate with other organizations to complete a project. GitHub encourages collaboration from the outside of an organization as a way of bringing new ideas and solutions and they consider it vital for the survival of an organization. The conductors of his experiment try to establish whichever exploration or exploitation is better for the overall progress of projects. In gathering and analyzing the data was used a web crawler powered by Python focused on GitHub projects bigger than 300 days and with at least five people. The number of commits was considered relevant for developers’ performance.

The results of the experiment show that successful repositories have more external developers but also a dedicated internal team that uses efficiently inside resources in their external interactions. In the cases followed with release software they also notice an increased external collaboration. As the researchers conclude, “we determined that the impact of exploration increases with an increase in exploitation, that is, ambidextrous research has a positive impact on the project performance of an open collaboration over the Internet.” (Lee, Baek, & Oh, 2020, p. 202). Three models reveal that a repository is successful if the number of external collaborations is higher and the performance drops if the number of internal members is higher. As Model 4 monitors software release cases, they found that performance is affected after the release just because all the development is switched to maintenance done by the internal team and external interaction is not mandatory anymore. This experiment demonstrates the importance of free unlimited interaction in OSSD. Exchange of ideas with collaborators outside the team proved to be beneficial for the success of the projects and for the future consistency of the teams.


Lee, S., Baek, H., & Oh, S. (2020). The role of openness in open collaboration: A focus on open-source software development projects. ETRI Journal, 42(2), 196–204.

1 thought on “Summary of Lee et al.’s “The Role of Openness in Open Collaboration: A Focus on Open‐Source Software Development Projects””

  1. TO: Professor Jason Ellis

    FROM: Nargis Anny

    DATE: September 23, 2020

    SUBJECT: 500-word summary Final Draft

    This is a 500 word summary of “A Smart Agent design for Cyber Security based on HoneyPot and Machine Learning”. The article highlights the rise of security risks that come with the rise of social media and the World Wide Web. We’re also introduced to the programs that keep the security programs running, as well as the setbacks it’s brings to computer systems worldwide.

    In the article, GDATA states how every year there are over millions of Cyber attacks that have been discovered. These issues are often involves analysis tools that keep track information. However, the difficulty is keeping an eye on every problem that arises. With a better understanding of how Cyber attacks work, there’s a better chance of preventing future issues. HoneyPots is one of the most prominent cyber security programs to date. Developed in 1992, HoneyPots is utilized as a monitoring and detecting system that locates harmful malware. Now future attacks can be prevented before they even find a system to disrupt. Part Two talks about Anomilies, data which has to be protected from harmful versions of software. With Social Media sites such as Myspace or Facebook, these sites need to be observed in order for a social ‘Honeypot”, to detect harmful profiles, as well as any other threats out there. Authors suggest a linkage defense system, which can bypass the setbacks brought on by past tools that tried to work. The Linkage system has the Honeypot’s and the defense system coexist together by having their management and communication tools work together. This system is based on a SMNP model code used in network management. Now Future intruders will be blocked by firewalls, if they try to hack into the system. Machine Learning is where we learn that computers operate under the system program that it’s been assigned. The concept of “Machine Learning”, keeps the computers adjusted to data structure and how to operate properly. Machine Learning has training models that separate into two phases in order to function. The first phase is estimating the data through training, by demonstrating tasks like recognizing animals in images or speech translation. The second phase is production. Here we see new data pass through the system in order to get the computer to complete an objective. The K-Means algorithm helps maintain clustering from certain systems. Eddabbah indicates that the “K –Algorithim is a faster solution to the issue it still has major setbacks” (Eddabbah, 2020, Page 3). The Decision tree helps branch out all data structures in case of testing. Part 4 jumps back into HoneyPot, this explains the different security communication networks. The first part is HoneyPot deployment which can monitor either Internal or External attacks on the system. With this we can see attacks that are carried out or attempted on any network. With DMZ’s (Demilitarized zones), HoneyPot function as a way to provide public internet service away from the computer’s internal network. Next, we have networks like KFSensor, Netfacade, Specter and CurrPorts. KFSensor is a server that watches out for connections with the network. Netfacade allows numerous network hosts interactions through unused IP a dresses. Networks also have to direct security threats to the firewall and eventually the honeypot will separate it to see if it’s serious or not. To conclude, network security is a very serious problem due to constant evolving and threats are hard to manage.


    Kamel, N / Eddabbah, M / Lmoumen, Y/ Touahni, R “A Smart Agent Design for Cyber Security Based on Honeypot and Machine Learning”, Security & Communication Networks, (2020) ID 8865474 (9 Pages), 2020

Leave a Reply