Empirical Speech and Language Processing

Language Selection

Breadcrumb Navigation


PhD students' report on conference visits in the USA


The PhD program offers financial support for conference attendance. Pinar Wennerberg (supervisor: Prof. Klaus Schulz) presented her phd project in computerlinguistics in USA at two conferences. Here you can read her enthusiastic report.

Report on conference visits towards obtaining a PhD degree in computational linguistics at CIS-LMU Munich

In March 2010 I attended two conferences and I gave an invited talk about my research activities related to my PhD thesis. The topic of my PhD thesis is Corpus Based Methods for Ontology Modularization in Life Sciences, which falls under the more generic field of knowledge management, yet with a specific focus on the clinical domain. For my thesis I have been investigating computational linguistics approaches to help identify small fragments of large terminologies. Such terminologies are commonly used in the clinical and biomedical domain, however as they are typically extremely comprehensive, they are not easily processable by software algorithms. Techniques for semi-automatic identification of relevant parts of the terminologies are therefore necessary.

This topic relates to various areas of computer science such as biomedical and clinical informatics and artificial intelligence. Consequently, I represented my latest research results at a conference related to the former i.e. Translational Bioinformatics Symposium of the American Medical Informatics Association and at one related to the latter i.e. Spring Symposium of the Association for Advancement of the Artificial Intelligence (AAAI).

The title of my first paper, which I presented in form of a poster, is Context Driven Modularization of a Representation of Human Anatomy. The second paper, which was similarly presented as a poster, is named Applications of an Ontology Engineering Methodology Accessing Linked Data for Medical Image Retrieval.

My poster received interest from a wide range of visitors, including professors, researchers and students. I made contacts with researchers from Stanford University, Columbia University, Mayo Clinic, National Cancer Institute etc. Some of the researchers have been facing similar problems or handling similar issues, therefore we decided to remain in contact. In general, I found it useful to make a first attempt to talk to the visitors when they are looking at the poster. Some of them may feel too shy to initiate the conversation. Therefore, I asked if they wanted me to explain what this was all about and almost all of them said yes.

The Spring Symposium of the AAAI is a traditional conference that has been held more than twenty years. It is well-known among the community and visited by members of academia (e.g. Stanford University, which is also the venue), industry (Microsoft, Yahoo, Google etc.), research organizations (PARC- Palo Alto Research Center, SRI International etc.). The symposium consists of several tracks on different topics. The track I attended to is called Linked Data Meets Artificial Intelligence. Linked Data is a newer concept relating to Semantic Web and various applications of it already exists across domains. Especially, around clinical and biomedical sciences there is an active Linked Data community. I presented my poster on the second evening, after the sessions on presentations. The community was smaller this time because it is a more focused research area. Here I made two interesting contacts. One is a researcher whose work I have been following. It was therefore nice to be able to meet him in person and to engage in a conversation face to face. He gave me useful information about other resources and communities and will remain in contact.

The second interesting contact is a researcher from PARC, which was previously the research and development department of XEROX.  They were just starting a project, which has overlaps with what I have been working on. After some conversation, I got invited to talk about my research at their centers. At the end of the conference, I indeed went to PARC and spent there about 2-3 hours discussing these matters and meeting the research team.

In between the two conferences I gave an invited talk at the Radiology and Information Sciences Department of the Stanford University. I gave an hour lecture on my research to the department that consists of faculty, students and additional visiting scientists. The talked was well received, and I spent the rest of the day there with follow-up meetings and discussions.

All in all this has been a very intensive period, which required a fair amount of energy, concentration and preparation. Nevertheless, I enjoyed every minute of it, and I am very happy that I was able to have the opportunity. I feel satisfied with the feedback I got. I also came back with new questions, which either directly came from the visitors or that emerged in my mind after discussions. Answers to these questions will certainly contribute to my thesis.

Pinar Wennerberg