Starting Lab Work


Author: Koji Sekiguchi

Several months into my school life at Japan Advanced Institute of Science and Technology (JAIST), I was assigned to a lab studying natural language processing (NLP). Meanwhile, a lecture on the theory of NLP has just begun and we have been studying how to learn corpus using various machine learning algorithm including Naive Bayes, decision tree, support vector machine, and hidden Markov model to do disambiguation of classification problem.

As I became curious about career path after graduate school for students who studied NLP, I asked my lab professor a question. Most of them “start their career as an SE at a major electronics company”, he said and added that they don’t utilize NLP in their business. As I went on to ask in detail, he said that some of the students join company laboratory specializing in NLP after doctoral course but students with masters degree only become “ordinary SE”. Evidently, even the professor had no idea how hot the technique we are using in his lab in some IT industries.

As the theme of professor’s work is highly difficult and the time when we will be able to process natural language in high accuracy would be still a long way off, It could be that he doesn’t have any idea that NLP could be utilized in the real-world in the first place. If the professor has that kind of feeling, graduating students will try to get a job in the real-world without the idea that their (fundamental knowledge of) theme is useful at work and end up “getting a job as an SE at major electronics company”. The professor’s phrase “getting a job as an SE at major electronics company” is a typical answer to a question asked by students and indicates that you can get a decent job even after doing researches on the field that we are specialized in.

I’m not saying that it is unfortunate to get a job at major electronics company to become an ordinary SE. You might get a job and work on a back-end system where I believe there are some area that you can apply your technique. I just thought that it’s a shame for the both academia and industry if you are not able to have a mindset that your fundamental knowledge acquired from your research can make the system you are working on more convenient as you get so used to the environment like this where you work on a back-end system.

» Pagetop