Skip to main content Skip to main navigation

Project | HyLAP

Duration:
Hybrid language processing technologies for a personal associative information access and management application

Hybrid language processing technologies for a personal associative information access and management application

In the project HyLaP, hybrid language and information processing methods will be improved, adapted and applied to a novel personal information access and management application. The application will facilitate the integrated access to both information in the user's own digital information space and on the global information structure of the WWW. Since we assume that language will remain the predominant medium for the storage and transfer of complex information including human knowledge, natural language processing methods will be key technologies for the exploitation of the abundance of information in daily work.

The resulting technology will combine open and closed domain question answering, named entity recognition, relation detection, automatic hyperlinking, and technology-assisted ontology building. Therefore, the research planned in HyLaP will continue the DFKI LT Lab's research and technology development in question answering (QA), information extraction (IE), natural language-enhanced information retrieval, and ontology-based knowledge extraction from human-language documents. This includes the participation in competitive mono- and crosslingual QA evaluation exercises. It also includes the ongoing search for effective and efficient hybrid language processing architectures.

The project will employ a combination of shallow and deep language technology methods for building a structured associative personal digital memory from a realistic large-volume collection of unstructured documents and structured data reflecting the digital information space of an individual user.

The goal is to demonstrate the virtues of such a structured memory together with improved IE and QA methods in an application that serves as intuitive and powerful associative interface between the user's thought and work on the one side and relevant related information in his personal memory and on the Internet on the other side.

The project has three main components:

  • Open domain question answering (ODQA)
  • Personal digital associative memory (PDAM)
  • Associative information access and management application (AIAMA)

which are established in two sub-projects:

  • HyLaP-QA for the research on open domain question answering
  • HyLaP-AM for the research on personal digital associative memory and application building.

HyLaP is funded under contract 01 IW F02.

Publications about the project

  1. Using Treebanking Discriminants as Parse Disambiguation Features

    Faisal Mahbub Chowdhury; Yi Zhang; Valia Kordoni

    In: Éric Villemonte de la Clergerie; Harry Bunt (Hrsg.). Proceedings of the 11th International Conference on Parsing Technologies 2009. International Conference on Parsing Technologies (IWPT-09), October 7-9, Paris, France, IWPT, 2009.

Sponsors

BMBF - Federal Ministry of Education and Research

BMBF - Federal Ministry of Education and Research