You managed to find me, Wouter Weerkamp... Maybe you're here by accident, or maybe you were actually looking for me. Either way, you're here now so I might as well tell you something about myself. For the visitors who just want to read one of my papers, here are my publications...

In February 2007 I started as Ph.D. student in the Information and Language Processing Systems (ILPS) group of the University of Amsterdam. We as ILPS conduct research in the fields of information retrieval, question-answering and natural language processing:

An information retrieval system does not inform (i.e. change the knowledge of) the user on the subject of his inquiry. It merely informs on the existence (or non-existence) and whereabouts of documents relating to his request ¹

Ok, that's your average introduction, but what is my position within this group of researchers? That, my dear visitor, is the issue of

Intelligent information access in user generated content

So how can we offer this "intelligent information access"? And why should we care about it anyway? Well, we can think of several characteristics that make user generated conent interesting:

• Noisy text: text generated by millions of people is certain to contain a huge amount of valuable information. Nevertheless, since it is user generated, it is much noisier than a more formal information source (e.g. news papers or encyclopedia). Retrieval models should take this into account and find ways of countering this (e.g. looking at credibility (pdf), or using external sources (pdf)) to offer access to the information in user generated content of offering access to information.
• Opinions: people tend to give their opinion on almost everything they can nowadays. Either they elaborate on these opinions in their blogs, they show their (dis)agreement with another blogger in the comments, or they respond to news events on news sites. Opinions are valuable, not only for people who want to create their own viewpoint, but also for companies ("what about my products?"), politicians ("what should my policy reflect?"), and intelligence ("what events could spark something"). Collecting opinions from user generated content is accessing the information in an intelligent way.
• Real-time: whenever something happens in the real world, people start writing about it online almost immediately. This real-time interaction between the real world and the online world creates nice possibilities, like estimating the importance of real world events, according to "normal" people instead of news editors, almost real-time! But we could also supply a way of archiving: summarize this day according to the internet users...

These topics are only a small part of what user generated content offers to the world of research. During my Ph.D. I will explore the topics above and possibly other topics that cross my path; I keep my options open. A quick look at the publications should give you more insight in what I've been doing.

More?

read (some of) my publications
ask me a question by sending an email
take a look at my running results
or expand you network on my LinkedIn profile

¹ LANCASTER, F.W., Information Retrieval Systems: Characteristics, Testing and Evaluation, Wiley, New York (1968).