Natural Language Processing and Digital Humanities is one of six groups that make up the Institute for Logic, Language and Computation at the University of Amsterdam. The group is focused on fundamental research into human information processing, especially in computational linguistics/natural language processing and digital humanities. It is a thriving research group, with 10 senior staff members and about 15 PhD students and postdocs. Members of the group have received VICI (2x), VIDI (5x), VENI (2x) and NWA-ORC (1x) awards from the Netherlands Organization for Scientific Research (NWO), and ERC Starting and Consolidator grants from the European Union. The group participates in the Language in Interaction Gravitation program of NWO, in Amsterdam Brain & Cognition and receives funding from a range of other organizations and companies (including SAP, Yandex, Google, Facebook and Amazon).
The group has a long history in those fields; since 1990, we have made many pioneering contributions to research on statistical parsing, dialogue modelling, information retrieval, machine translation, the evolution of language and music, and deep learning.
- Our online statistical parsing demo, intgerating much of our work in Data-oriented parsing and applied to English, French, German & Dutch, is still available here: demo
In our work in Computational Linguistics we develop state of the art statistical and deep learning models (for machine translation, dialogue, computational semantics, parsing and other domains), while incorporating insights from linguistics and cognitive science. The ILLC’s computational linguistics meet regularly (since 2004) at the Computational Linguistics Seminar, which features top researchers from around the world. They also organize various other lab meetings and reading groups, including the deep learning reading group (since Jan 2016).
In our work in Digital Humanities, one important application area is Information Retrieval and Question Answering. In this area, we employ the state-of-the-art retrieval techniques, and focus on improving the practical usefulness of these systems through innovations in user-interfaces and cognitive ergonomics.
- Many Digital Humanities researchers at ILLC are part of the Department of Media Studies.
The methods we use in Language & Computation thus range from mathematical and computational techniques (including symbolic, statistical and neural network models) to experimental methods in collaboration with linguists, psychologists and cognitive neuroscientists.