Scholarly Communication

Listen to this interview of Lianglu Pan, PhD Student, and Shaanan Cohney, Senior Lecturer, and also Thuan Pham, Senior Lecturer — everyone at University of Melbourne, Australia. We talk about their coauthored paper EDEFuzz: A Web API Fuzzer for Excessive Data Exposures (ICSE 2024).Thuan Pham : "The reading pattern in our group goes something like this: When reading to broaden our knowledge and come up with ideas, we focus on the conceptual contribution of a paper, instead of zeroing right in on the technical side. Because, when the conceptual side is good, then the paper can be readily applied to similar problems — and what's more, the technical side becomes vastly easier to understand once you've understood the concept to begin."Writing guidance mentioned in the episode: Chicago Writing Program and Joseph William's book Style Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Lorenzo Rossi, Research Fellow, University of Camerino, Italy. We talk about his coauthored paper A Technique for Discovering BPMN Collaboration Diagrams (SOSYM 2024).Lorenzo Rossi : "Yeah, this way of structuring the concluding remarks in this paper is a technique we often apply in our research contributions, especially to journals, where the space limitations are less stringent. This structured approach to the conclusion, where we discuss assumptions and limitations as here, really should almost be somehow mandatory for good research work. Because, especially in foundational research, there's no way around making certain initial assumptions. So, without writing these into your published work, you are (consciously or unconsciously) hiding weaknesses of your contribution, and therefore making it harder for others to build upon your findings." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Kangfeng Ye, Research Associate, University of York, UK. We talk about his coauthored paper Probabilistic Modelling and Verification Using RoboChart and PRISM (SOSYM 2022).Kangfeng Ye : "In this paper, I have four coauthors, all of them senior researchers. And when we reviewed the manuscript internally, we adopted a strategy we call sequential review. In the usual process of review at a conference or journal, every submission gets reviewed simultaneously — all reviewers receiving the same manuscript at the same time. However, we ran our internal review (that is, our revisions before submission) in a sequential fashion: I provided the first draft to one coauthor for review, they gave their feedback, I revised in order to provide that next draft to a different coauthor for review, and so on." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Emerson Murphy-Hill, Research Scientist, Microsoft. We talk about his coauthored paper GenderMag Improves Discoverability in the Field, Especially for Women (ICSE 2024).Emerson Murphy-Hill : "Too often in papers, the authors get defensive about limitations or threats to validity. Of course, they'll state outright a limitation, like in our paper that we study only one small feature of a company-internal piece of software. But many authors will then grow defensive, claiming, like, 'Well, this is actually a really important piece of software and it's used by tens of thousands of users — our numbers are really big!' But I don't really think that that resonates with readers. I think the defensiveness comes across pretty transparently. So, I think just addressing things head-on is a more effective strategy for having a good and honest conversation with readers and with reviewers." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Ionut Predoaia, Research Fellow, and also, Antonio García-Domínguez, Senior Lecturer — both at the University of York, UK. We talk about their coauthored paper Streamlining the Development of Hybrid Graphical-Textual Model Editors for Domain-Specific Languages (ECMFA 2023).Antonio García-Domínguez : "I think that the limitations in any work are really opportunities for follow-up research. I mean, essentially, you are identifying for the reader, 'Look, these are the bits that we've not handled just yet — and obviously, we will likely be the first ones to try to tackle that' — but, you know, there's no reason why really any other researcher in the community wouldn't attempt to tackle that from their angle or for their research purposes. They may have the better idea even, right." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Zejun Zhang, Research Scientist, Nanyang Technological University, Singapore. We talk about her coauthored paper Hard to Read and Understand Pythonic Idioms? DeIdiom and Explain Them in Non-Idiomatic Equivalent Code (ICSE 2024).Zejun Zhang : "Following my presentation of the paper at ICSE, it was interesting. I mean, there was, first off, a lot of positive response, but then some people in the audience were asking why we would research the readability of Pythonic idioms, and also, why we would translate those idioms into non-idiomatic code. Now, these questions were coming in relation to our previous work on idiomatic code. Nonetheless, the effect for me was that, for future work, we need to further explore this line of the research and really explain Pythonic idioms so that developers can deeply understand them." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Roberto Verdecchia, Assistant Professor, University of Florence, Italy; and also, Per Runeson, Professor, Lund University, Sweden. We talk about their coauthored papers Threats to Validity in Software Engineering Research: A Critical Reflection (IST 2023) and Threats to Validity in Software Engineering — hypocritical paper section or essential analysis? (ESEM 2024).Per Runeson : "I think what we've seen in our work here on threats to validity — and it was certainly our intention in conducting it in the first place — is, to have the researcher take the initiative and really adopt a reflective attitude. Because, research is not only about investigating facts and testing hypotheses. It's also about reflecting on the learning process, and it's about the extent to which you can trust what you've learned in doing that research, but also it's about the way forward from there, that is, how do we take the work forward into the future." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Nan Jiang, PhD candidate, and Lin Tan, Professor — both at Purdue University. We talk about their coauthored paper Impact of Code Language Models on Automated Program Repair (ICSE 2023).Lin Tan : "In my research group, the procedure in every project is to write the Introduction early — very early, in fact. It's the first section I have my researchers think about, actually. Because, you know, a lot of people will imagine that the approach section is where you begin — basically, to write exactly what it is that you did. But the advantage of beginning at the Introduction is that you clarify the contributions, you define the problem and also understand well your reason for tackling it. So, typically some three months before the deadline, I have my researchers really start sketching the Introduction in the manuscript." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Jenny Liang, PhD student, Carnegie Mellon University. We talk about her coauthored paper A Qualitative Study on the Implementation Design Decisions of Developers (ICSE 2023).Jenny Liang : "When it comes to selecting specific results or codes, I like to think about it in terms of what was surprising. So, maybe it's not so surprising that people think about requirements when making these implementation design decisions — and that's why we didn't talk about that. But what will be interesting, for example, is the fact that they think about future requirements that might come down the pipeline — and so, that's why we selected that. So, that is one heuristic, for me, basically: know what the prior literature is, know what the relevant community believe — and then cater to that." Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen to this interview of Yun Peng, Research Associate, Chinese University of Hong Kong, China; and also, Cuiyun Gao, Associate Professor, Harbin Institute of Technology, China. We talk about their coauthored paper Static Inference Meets Deep Learning: A Hybrid Type Inference Approach for Python (ICSE 2022).Yun Peng : "And I remember the reviewers at ICSE commenting how they never imagined solving the type-inference problem in just this way. So, for me, the takeaway here, is: When we are conducting research or writing a paper or solving a technical problem, we do well to look into life and draw inspiration from there to do the work — because I know we can be greatly inspired by the things just around us in our everyday lives." Learn more about your ad choices. Visit megaphone.fm/adchoices