The Information and Language Processing Systems group is part of the Informatics Institute of the University of Amsterdam. Our research is aimed at intelligent information access, especially in the face of massive amounts of information. We work on finding and analyzing content (information retrieval, machine translation, language technology), the analysis of structural information (social networks, linked data) and the analysis of user behavior (self-learning search, log analysis, user studies).

We combine fundamental, experimental and applied research, and we do so using a broad range of textual data, data from the web or enterprises, edited or user generated, or obtained from (automatic) transcriptions of audio or video. We are involved in a large number of projects with other groups, both within and outside academia. Our research is funded by NWO, KNAW, the EU and through a range of public-private partnerships.

Recent News

Dataset Release for Evaluation of Counterfactual Algorithms

Criteo Research is pleased to announce the release of a new dataset to serve as a large-scale standardized test-bed for the evaluation of counterfactual learning methods. Criteo Research has access to several large-scale, real-world datasets that they would like to share with the external research community with the goal of both advancing research and facilitating an easier exchange of ideas. The dataset they are releasing has been prepared in partnership with Cornell University (Thorsten Joachim’s group) and the University of Amsterdam (ILPS). Continue reading

FAT/WEB: Workshop on Fairness, Accountability, and Transparency on the Web

www2017_logoRecent academic and journalistic reviews of online web services have revealed that many systems exhibit subtle biases reflecting historic discrimination. Examples include racial and gender bias in search advertising, image recognition services, sharing economy mechanisms, pricing, and web-based delivery. The list of production systems exhibiting biases continues to grow and may be endemic to the way models are trained and the data used. Continue reading