This post is from Roman Haas (he was advised by Elmar in his Bachelor’s Thesis).
I presented my paper about “Deriving Extract Method Refactoring Suggestions for Long Methods” which I extracted from my bachelor’s thesis as part of the Software Quality Days 2016 conference. This post describes how I ended up at the Software Quality Days, how I prepared my presentation, what I didn’t know about conferences so far, and what I want to do differently in my next presentation.
Background
My bachelor’s thesis was about “Deriving Refactoring Suggestions from Quality Analysis Findings”. I analyzed a couple of software quality analysis finding classes and common countermeasures. Some classes of findings can be fixed trivially (e.g. missing parenthesis like ‘{‘, ‘}’ that can just be added). For other findings – in general – several “easy” solutions exist. But depending on the system context only one is appropriate (for example “non-constant public variable” or “unused local variable” findings). For a third group of findings there are no easy solutions – like code clones or long methods. For each individual finding, a developer needs to find an appropriate solution by considering several alternatives. The main part of my bachelor’s thesis was about automatically deriving extract method refactoring suggestions for long methods.
Recycling Results from My Thesis
When I had finished my bachelor’s thesis, Elmar asked me whether I would like to publish my results in a scientific paper. He suggested the Software Quality Days (SWQD) – a business conference with a scientific track – as a suitable conference. The topic of my paper fitted well to the topics covered on the SWQD. The conference takes place in Vienna every year and is not that big (300-400 participants on 5-6 tracks), which was advantageous for me as it probably increased the chance to get the paper accepted and was a good opportunity to gather first conference experiences.
From Bachelor’s Thesis to Scientific Paper
Preparing the paper was quite straight-forward: I had all the results, and as I had written my thesis in English I was able to recycle parts of it.
The image below gives an overview about my paper and how much rework was necessary for the single paragraphs. Red paragraphs are completely new, yellow ones were rewritten but their contents are similar in my bachelor’s thesis and green paragraphs needed no or only minor changes.
The introduction is the only chapter that I had to redesign from scratch. That is because the paper is much more focused on a single topic and needs to be motivated in another way than the more general topic of my bachelor’s thesis.
I was able recycle big parts of the Approach chapter. Of course, I had to shorten it a lot but there were only a few paragraphs that needed to be changed to keep the text comprehendible.
Even though I could recycle a lot of text, it took several days to extract the paper from my bachelor’s thesis, especially because many small changes needed to be applied to ensure coherence within the text.
Transforming the Thesis Presentation into a Conference Presentation
I was a bit naive when I thought that recycling the presentation of the thesis was as simple as recycling the written work. Even though the bachelor’s thesis presentation was well-prepared and I had presented it already several times (in German and English) I had shorten it by five minutes; instead of twenty minutes as in the previous instances of the talk, I was only allowed to talk for fifteen minutes at SWQD. My first idea was just to shorten the presentation by deleting some slides.
Rehearsal
The rehearsal (done with my co-author and thesis supervisor) revealed that some important parts were not clear enough and my audience could not follow all my explanations. In addition, I got the advice to focus even more on the contribution of the paper. A major difference between the presentation of the bachelor’s thesis and the paper is the focus of the presentation. Whereas you usually mention all important parts and steps of your thesis shortly in the first case, you focus much more on your contribution (your new ideas) in the second.
Additionally, my co-author (advisor of the thesis) pointed out that one should avoid the usage of the laser pointer as it may happen that the screen is positioned in a way such that the presenter cannot easily point on it. Therefore, I added highlighting elements to my slides; so I had not to rely on the unknown equipment of the conference.
Spending some time on a rehearsal is a very good investment and I highly recommend it to everyone who wants to give nice presentations.
How Did the Presentation Turn Out?
My presentation was scheduled for the morning of the second conference day. This time slot had a big advantage: I had a day to acclimatize at the conference and to get a feeling how presentations work at scientific conferences. I realized that presentations at the university differ not that much from conference presentations in general. There was a wide range of topics covered by the presented papers and the quality of the presentations varied a lot.
All presentations of the scientific track took place in the same room, a relatively small one with seats for up to 40 people. The size of the audience varied and was between 5 and nearly 40 participants. Because of the rather small room the presentation setting was similar to the setting of university presentations – no large distance between you and the audience, screen and laptop directly next to you, and no need of a microphone. About 20 people listened to my presentation; a quite good number that fit perfectly into the room.
For each conference session (time slot between the coffee breaks) there was a moderator who announced the presentations at the beginning and moderated the discussion afterwards. Unfortunately, there were not always questions from the audience. Depending on the moderator, he or she asks a question himself (which then usually leads to further questions from the audience) or the session continues with the next presentation. I was quite happy that I received many interesting questions immediately after the presentation.
Noteworthy About Conferences
Usually it is possible to use your own equipment for the presentation. In my opinion it is a good idea to take this possibility. If you don’t, Murphy’s law is likely to catch you as the following examples show: During a keynote the presenter (technical device) ran out of battery and the presenter (human) had to switch slides by clicking on the touchpad. But if you use your own equipment, be sure that everything is up and running (and survives your presentation). One presenter who used his own laptop forgot to charge the battery, leading to a black screen during his presentation. He had to nervously search for his power supply. Obviously, this destroys unnecessarily the flow of the presentation and additionally costs time and nerves.
It is a good idea to inform yourself about the dress code of the conference. On the SWQD there were many exhibitors (who of course wore suits), but on the scientific track jeans and shirt were fine. For computer scientists, this will probably be okay for nearly all scientific conferences.
A colleague of mine gave me a good advice: ask one or more questions in the discussion session of the presentation right before your own presentation. This makes your speech engines warm up, reduces your nervousness and motivates your audience to ask questions after your presentation.
What I Want to Do Differently Next Time
- Often, the moderator announces the presentation with the name of the presenter and the title of the presentation (the title of the corresponding paper). To avoid annoying duplications, I want to prepare an additional introduction to my presentation that does not contain the title. However, it makes sense to have an alternative beginning that contains the title if the moderator does not mention it.
- I did further research on the topic, but presented only the parts covered in the paper on the conference. Next time, I will give an overview about the current state of my research, even if I did not publish yet the corresponding results.
- In my presentation, I focused for the evaluation part on the results and not the setting of the evaluation (I explained the setting roughly, but had no slide containing details on the evaluation setting). In the discussion afterwards it became clear that this was a bit too short. For my next presentation I will include a slide that visualizes the key values of the evaluation setting (like the number of developers that I asked and their experience).
- There was a question about the usefulness of the suggestions of the approach. I argued with my tiny evaluation (which is based on hard numbers but has only a very small sample size), but did not mention that I implemented the approach in Teamscale (a static analysis tool that is developed at CQSE). The conference guidelines claim that no advertisement may take place on the scientific track and therefore I did not want to mention the commercial product. But to convince the audience that the approach works, it would have been perfectly okay to talk about the experiences with the refactoring suggestions in Teamscale.
- At the evening before my presentation I went to sauna and did not drink enough afterwards, which resulted in a big headache the next day. Remember: Always drink a lot after going to sauna, especially if you have to give a presentation the next day.
- On the next conference, I will make better use of the break before my presentation (talking is okay, but do not forget to go to the restroom, especially if you had to drink a lot to get rid of the sauna-headache).
Conclusion
Presenting my paper at the Software Quality Days was a great experience. The conference is optimal for beginners because you can practice to give conference presentations with a smaller (but not less interested) audience. I enjoyed my paper project and would like to do it again. As already mentioned, I did further research on the topic and with a bit of luck I may present another paper on a scientific conference, soon!
Thesis Guide 