Current research interests
The questions I address in my research are
- How do people understand what they read?
- How do authors' intentions emerge in the text?
... which two questions I consider to a large extent to be the same. I have pursued this study through three sets of research tasks:
- The study of how topical structure and textual style is transmitted through markers in text. This I have done through a number of quantitative studies of text, usually evaluated in terms either of statistical hypothesis testing or in terms of usefulness for information retrieval tasks.
- The study of how text is understood, assessed, and -- tentatively -- how it is trusted by its readers (in cooperation with my colleague Preben Hansen). This we have done through a number of user studies formulated as information access situations.
- The study of how a data-oriented model of semantics can be built from analysis of text (in cooperation with my colleague Magnus Sahlgren). This we have done through a large number of quantitative studies of text using a novel framework first proposed by Pentti Kanerva; the results are evaluated in terms of their usefulness for information retrieval tasks.
Background
I received my B A in Computational Linguistics and Mathematics in 1988, a Licentiate of Philosophy in Computer and Systems Sciences 1992, and a Ph D in Computational Linguistics in 2000 - all at Stockholm university. In 2006 I was appointed Docent (adjunct professor) of Language Technology at the University of Helsinki.
I have since 1990 been employed at SICS, the Swedish Institute of Computer Science, in various instantiations of the Language and Interaction laboratory. I have previously worked at SISU, the Swedish Institute for System Development, been a visiting student at Columbia University, a programming assistant at Xerox PARC, and an assistant research scientist in the PROTEUS project at New York University. During the academic years 97-99 I served as professor of computational linguistics at Helsinki University; I have also taught and supervised students at Stockholm University and at the Royal Institute of Technology in Stockholm, participated in numerous research projects, national and international.
I frequently serve as a member of the Program Committee for the SIGIR conference, and have organized and (co-)chaired a number of recent workshops on broadening the scope of information retrieval.
Recent Workshops and Symposia
- 2002 - AAAI Spring Symposium on Acquiring (and Using) Linguistic (and World) Knowledge for Information Access
- 2005 - SIGIR workshop on Stylistic Analysis of Text for Information Access
- 2006 - EACL workshop on New Text - Wikis and blogs and other dynamic text sources
- 2006 - SIGIR workshop on Stylistics for Text Retrieval in Practice
Undervisning
Verktyg - Tools
Simple Document Assessment tool for CLEF
Ickeparametrisk hypotesprövning - Non-parametric hypothesis testing
- Mann Whitney U (Test to see if some category tends towards high or low measurement values (equivalent to Wilcoxon)).
- Spearman rho (Test to see if two measurements vary together - rank sum correlation).
- Khi 2 (Test to see if two categorial distinctions have dependencies).
- Kruskal-Wallis (Multi-variable Mann Whitney U).
Textstatistik - Text statistics
Julklappsprogramvara
Publikationslista - List of publications
- An Non-Exhaustive List of Publications, Many Available On-Line in Various Variously Quite Popular and Portable Formats
- Research Notes - Unpublished Ideas, Starting points, etc etc