Organisation profile: Natural Language Processing Research group - University of Leeds

---

Research areas

Leeds has a long and well-established record in natural language and
speech processing. Research during the last four years has focussed on
the following areas:
  + Work on spoken dialogue processing has focussed on extending
    language understanding systems to include context-sensitive
    dialogue information, such as speech-acts, topic-focus and
    co-reference; and the development of evaluation procedures for
    dialogue management systems.
  + In the field of speech recognition, an EPSRC Advanced Research
    Fellowship was granted which is devoted to combining linguistic
    and statistical information to arrive at improved language models
    for speech. Cooperation with Xerox PARC (Natural Language Theory
    and Technology group) has resulted in a linguistically motivated
    language model for speech processing using the data-oriented
    parsing technique combined with Lexical-Functional Grammar
    representations.
  + The EPSRC-funded AMALGAM project has been especially successful in
    creating mappings among lexico-grammatical annotation formalisms,
    and has led to the development of part-of-speech taggers for
    virtually all state-of-the-art annotation schemes.
  + Work on Language Engineering applied to English Language Teaching,
    including collaborative EU project ISLE: Interactive Spoken
    Language Education, has been highly regarded by ELT practitioners:
    the British Council commissioned and published an overview to
    promote British excellence in this field worldwide; we provided
    invited speaker at ITI99 (Institute of Translators and
    Interpreters), IATEFL99 (International Association of Teachers of
    English as a Foreign Language), TESL-Canada2000 (Teachers of
    English as a Second Language), as well as reaching a Language
    Engineering audience.
  + Developing research directions in NLP include the detection of
    intelligent language-like features for SETI (Search for Extra
    Terrestrial Intelligence); the extension of the data-oriented
    parsing model to RNA parsing, the annotation of language corpora
    with speech-act information, the bootstrapping of syntactic
    structure for unsupervised language learning and automated
    abstracting for document retrieval.