An Empirical Study of Chinese Name matching and Applications

Introduction
1. Name matching
  1. Important
    1. Downstream tasks
      1. Entity linking
        Includes context of mentions
      2. Entity clustering
        Includes context of mentions
      3. ?
      4. Entity coreference
        ?
      5. Name transliteration
      6. Identifying names for mining paraphrases
        ?
      7. Standalone name matching
        Context independen
    2. Entity disambiguation
      1. Determine if two mentioned strings refer to the same entity
  2. Methods
    1. Language type
      1. Alphabetic languages
        Focused on
        Example
        English
        Indo-European
      2. Logogram languages
        Example
        Chinese
        Hanzi
        Challenge
        A small number of hanzi represents an entire name
        There are X*10.000 hanzi in use
        Current methods
        Largely UNTESTED
        Coreference resolution errors
        Caused by
        Chinese name matching errors
    2. Focus on persons names
  3. Challenge
    1. Issue: Name variations
      1. Nicknames
      2. Aliases
      3. Acronyms
      4. Differences in translation
    2. Exact string matching
      1. POOR results!
  4. Determine whether two strings refer to the same entity based on the strings above.
Research
1. Evaluate Name Matching methods
  1. In Chineese
  2. Approaches
    1. Existing
      1. String matching
        ?
      2. Learnig
        ?
    2. New
      1. New Representation for Chinese
  3. Experiments
    1. New Representation for Chinese
      1. Improves
        name matching
        Entity clustering
      2. No details?!
2. Newly developed data sets
  1. Matched Chinese name pairs
3. Mingpipe
  1. Name matching tool
    1. Python package
      1. Usage
        As stand alone
        Integrated in a larger system

Siguiente

An Empirical Study of Chinese Name matching and Applications

Descripción

Resumen del Recurso

Similar

	Creado por Ivan Zapreev hace más de 8 años