I have started an Interview category post on my blog so that you can learn something useful from interviews with researchers. I had a chat with Dr. Philippe Fournier-Viger to obtain information about computer science research and other related activities. Before commencing the chat, I’d like to tell you about him and his research blog.
Dr. Philippe Fournier-Viger is a great data mining researcher and Assistant Professor, working at a reputed institution. He is an inspiration for budding researchers. He is the founder of the SPMF data mining library. The library offers 110 data mining algorithms and can be used for further research work. The spmf library has been cited in more than 200 papers. He has published more than 80 research papers, mostly related to data mining.
Readers can visit the following links to know more about him and its data mining library.
Hi Dr. Philippe,
1: You have done many researches in the field of Computer Science and founded an open source library that can be used by the budding researchers for further development. This quality differentiates you from other researchers. What is the reason behind it?
Hi Rahaman. Thanks for inviting me to this interview. I have started to work on the SPMF library during my Ph.D. The main motivation was that I had observed that many data mining researchers do not publish their source code and datasets. When researchers do not publish their source code and datasets, other researchers who want to use their work need to implement the algorithms again. Thus, these researchers spend a lot of time implementing again the work of others instead of working on their own research. For this reason, I have founded SPMF. The goal is to provide a wide array of algorithms for data mining and in particular, pattern mining, so that other researchers can use these algorithms in their research without having to implement them again. Also, when I created SPMF, I wanted to give a good example to other researchers and I hoped that other researchers would then do the same.
The first version of the library was simply a small project that I did for a data mining course. I had implemented about five algorithms. Then slowly, I have implemented more algorithms over the years, either for my research or for my own curiosity. Then, also several contributors have provided source code to be included in the library.
In my opinion, each researcher should try to publish its source code and datasets, when it is possible. There are several reasons. First, it facilitates the work of other researchers. Second, it also increases the possibility that one’s work is reused by others, and thus cited by other researchers.
2: You could have join in IT companies, why did you choose the teaching field.
During my bachelor degree, I decided to do a master degree and Ph.D. to see how far I could go. It was a kind of challenge that I set to myself. I wanted to see if I would be able to do the Ph.D.
A first reason why I choose to be a professor after the Ph.D. instead of working in the industry is that I enjoy the freedom that I have in academia, when doing my research. As a professor, we have a lot of freedom about the research projects that we want to do. I can work on designing algorithms that are applied but also on more theoretical projects. In a company, I would have more to work on applied projects and I would certainly have less choice about the projects that I could work on. Also, I really appreciate to have the freedom to publish the source code of my research rather than working on something that belong to a company.
Besides, I really enjoy the schedule of a professor. Although a professor has a lot of work, especially during the first years of being a professor, an advantage is that the schedule of a professor can be very flexible. At the university where I was previously working, I had to be at university about 10 to 15 hours a week. Then, I could also work from home.
I also enjoy teaching Every time that I teach a new course, I need to read some new books and I also learn new topics at the same time that I’m teaching them.
Another reason is job security. To get a professor job can be very competitive in some countries. But once one gets a tenure-track position, one can enjoy job security.
Another thing that I like about being a professor is the opportunity to travel to attend international conference with other researchers. Since I like travelling, this is something very interesting for me.
3: What are the skills need to be developed for a student to become a successful researcher like you?
To be a successful researcher, I think that being able to work hard day after day and during many hours is one of the most important skill. Also, one needs to like the research that he is doing. Otherwise, it is easy to lose the motivation and quit, especially during the Ph.D. Personally, I work very hard. During my studies as M.Sc. or Ph.D student, I often only took a few days of vacations every year. But I think that it is also important to keep some balance and find time to do some sport and eat well, to stay healthy.
Besides that, to be successful in computer science, one should be good at programming. This is a skill that can be acquired over time by practicing. Also, if one is good at mathematics, it is certainly an advantage for several topics in computer science.
To be successful in research, one should also try to work with other researchers instead of just working by himself, and create collaborations, especially during the Ph.D. For example, during my Ph.D., I often worked with other students, and thus I had several publications with other students, which increased my total number of publications. Also, after the Ph.D., as a professor or researcher, it is important to have relationships with other researchers. Sometimes, it leads to more opportunities.
Another important skill to learn for a young researcher is to write well. For non-native English speakers, it is important to learn to write well in English because most of the research is published in English nowadays. Besides, it is important to learn how to write scientific articles. This is something that takes time to learn, because it is learned by practice (by writing articles). The first article is the most difficult to write. Then, the more you write articles, the more it becomes easier.
4: I have read many of your articles and astonished the way of presenting the article. How to improve the quality of writing?
Thanks! Writing is a skill that is acquired by practice. My first article was terrible. To get better at writing, it is necessary to write many articles and not write them in a rush, but take the time to read them again and again, and try to improve them. The more you read articles, the more you also understand how articles should be written.
There are at least two key challenges when writing a paper: (1) writing well in English and (2) how to organize the content of the paper.
A good way to learn how to organize the content of a paper and to get more familiar with the writing style is to look at some papers that are well-written in top journals or conference and then to try to follow the same writing style when writing your paper.
Other ways to improve writing is to ask help from a senior researcher or students. For example, a master degree student could ask a Ph.D student to give his opinion on the paper written by the master degree student.
5: Fake researches, journals and plagiarism are the problems revolving around the world. What will be your valuable opinion to sort out the problems?
Yes. Those are important problems. For plagiarism, I think that the main solution is that institutions would give more severe punishments. For example, recently, I found that some researchers from a university in India had plagiarized some parts of one of my papers. I thus sent an e-mail to their institution to let them know about this and also to the journal where the paper was published. The paper was taken down from the journal. However, there was to my knowledge, no punishment at all from the institution toward the researchers who had plagiarized my paper. If the researchers who plagiarized where punished more severely, others would be less likely to plagiarize.
For the young researchers, I highly recommend to avoid plagiarism. It is something that can be very bad for one’s career. Moreover, nowadays, it is very easy to detect plagiarism using search engines.
For the problem of fake journals, I think that young researchers need to know more about how to distinguish fake journals from good journals. Institutions and professors should explain to their students how to choose a good journal/conference. Moreover, institutions or government should have some ranking systems to evaluate whether a journal or conference is good or not. In some countries people rely on some metrics such as the impact factor or whether the journal is indexed in SCI, EI, and other indexes. This is some way of avoiding the problem of fake journals.
6: Express your insights on Data mining and suggest some other fields for the new researchers?
Data mining is a very popular field currently. I think that it is a very promising field in computer science. To choose a good research topic in data mining, I would recommend to look at the papers published in top data mining conferences and journals, during the last one or two years. Currently, there are several trends such as big data, social networks and deep learning. I think that one should choose something that is popular but also something that he likes. Moreover, one may also choose a topic based on his skills. For example, a student that is good at math may choose a topic that require more advanced mathematics.
7: My last Question, “Simplicity is the reason for the success of Dr.Philippe”, am I right?
Interesting question! I think that there are several important factors for being successful, as we discussed previously (working hard, choosing good topics, collaborating with other researchers, writing well, etc.). Simplicity can be viewed from different perspectives. For a teacher, it is important to be able to provide simple explanations to students. For a researcher, it is important to express his idea and describe his research in a way that is easy to understand by other researchers. Learning to explain simple or complicated ideas in simple terms is thus a key skills for researchers.
“I sincerely thank you for spending your valuable time for my blog readers. On behalf of readers, I wish you to do more successful research to help the computer science community. ” – Rahaman
“Thank you very much. It was my pleasure to participate to this interview. “