• DocumentCode
    3300766
  • Title

    Toponym Resolution in discourse

  • Author

    Tang, Xuri ; Chen, Xiaohe ; PENG, Minxuan

  • Author_Institution
    Sch. of Language & Literature, Nanjing Normal Univ., Nanjing
  • fYear
    2008
  • fDate
    19-22 Oct. 2008
  • Firstpage
    1
  • Lastpage
    8
  • Abstract
    Toponym resolution involves two stages: toponym recognition and reference resolution. The present research proposes an integrative model for the resolution of Chinese toponym reference in discourse. The model employs the concept of cognitive salience, an inherent property of toponym references, to integrate different heuristic reference disambiguation methods into a coherent design. The heuristic methods involved in the model are category salience, geographic distance salience, co-occurrence frequency salience and discourse distance salience. An algorithm is also designed and implemented to evaluate the performance of the model. The evaluation is focused on the resolution of Chinese toponyms within the linguistic unit of discourse. Experiments of evaluation report satisfactory results with a best precision of 91.1% and a best recall of 73.6% for reference resolution.
  • Keywords
    linguistics; natural language processing; Chinese toponym reference resolution; category salience; co-occurrence frequency salience; cognitive salience; coherent design; discourse distance salience; geographic distance salience; heuristic reference disambiguation methods; integrative model; linguistic unit; toponym recognition; Algorithm design and analysis; Business; Character recognition; Frequency; Geographic Information Systems; Grounding; Information analysis; Natural languages; Spatial resolution; Springs; Cognitive salience; discourse; reference disambiguation; toponym matching; toponym resolution;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Natural Language Processing and Knowledge Engineering, 2008. NLP-KE '08. International Conference on
  • Conference_Location
    Beijing
  • Print_ISBN
    978-1-4244-4515-8
  • Electronic_ISBN
    978-1-4244-2780-2
  • Type

    conf

  • DOI
    10.1109/NLPKE.2008.4906777
  • Filename
    4906777