Internship: “A Win-Win Process to the University, to the Company, and to the Student”
Originally from Lisbon, Portugal, André Torres Martins is a dual degree Ph.D. student in Language Technologies (LTI) at the Instituto Superior Técnico of the Universidade Técnica de Lisboa (IST/UTL) and at Carnegie Mellon’s School of Computer Science. Prior to joining the doctoral program, Martins worked in research and development in computer vision and image processing at Reverse Engineering, and in research and development in language technologies at Priberam Informatica. Currently, André Martins is carrying out a three months internship at the Google Research in New York.
In the U.S. it is common for the doctoral students to complete internships while taking courses. The goal is to immerse students in the industrial world. For André Martins “this kind of paid internships should be put into practice in Portugal to ease the link between academia and companies,” adding that “at the same time it would give the student the opportunity to confront with real problems that occur inside the companies.” This doctoral student feels that “this could be a win-win process to the university, to the company, and to the student.”
Being part of the natural language processing research team at Google, André Martins is helping to develop algorithms for large scale semi-supervised learning using Google’s infrastructure, the aim being to improve text analysis by using large amounts of data collected from the web. This poses new challenges that have not been much studied by the academic community, since universities do not usually have access to such massive data centers.
From Portugal to the U.S.
André Martins had already spent two years at Carnegie Mellon University, in Pittsburgh, so for him it was not complicated to get back to the U.S.. “This has been an amazing personal and professional experience, and I feel totally part of the research team.” For him “the environment inside Google is similar to the one that we feel in a top university, but with a big emphasis in engineering and applied research in large scale problems.” At the professional level, André Martins says that there are differences related to the productivity if we compare Google with some Portuguese companies. “Productivity doesn’t mean to work more hours, but it is a result from a different way of managing time, a better division of tasks, a better management of the resources, among other issues,” adding that the “American industrial culture is also very different: they give more importance to the human resources qualification and to the level of fulfillment and motivation of the worker."
André Martins research interests overall include machine learning and natural language processing. Martins decided to come to the dual-degree program because he felt it was a good opportunity to enhance his skills in language technologies, to make research on many exciting open problems, and to interact with some of the best researchers in the field. He hopes to continue doing research on language technologies as his career develops. During his dual degree Ph.D., through the Carnegie Mellon Portugal program, he is advised by Mário Figueiredo (IST/UTL), Pedro Aguiar (IST/UTL), Noah Smith (CMU), and Eric Xing (CMU). He earned his undergraduate degree in Electrical and Computer Engineering from the Instituto Superior Técnico of the Universidade Técnica de Lisboa.