Fabien Duchateau

Page de Fabien Duchateau

Site web UCBL Site web LIRIS


Version anglaise

KIEV

KIEV (a.k.a. SPIDER) aims at extracting binary relationships from textual documents to populate semantic triple stores or large knowledge bases. It combines a semantic part (extension of labels, clustering of frequent terms) with natural language processing techniques (Part Of Speech tagging) to generate relevant patterns for a specific type of relationship. Three use cases are presented: the former discovers the type(s) of relationship between two given entities (e.g., 'parody' between entities 'lord of the rings' and 'bored of the rings'). The second use case finds all related object entities given an initial entity and a type of relationship (e.g., object 'lord of the rings' for 'bored of the rings' + 'parody'). The latter discovers examples (i.e., pairs of entities) which respect a given type of relationship.

Related publications

  1. KIEV: a Tool for Extracting Semantic Relations from the World Wide Web
    International Conference on Extending Database Technology (EDBT), 2014
    Naimdjon Takhirov and Fabien Duchateau and Trond Aalberg and Ingeborg Solvberg

    @inproceedings {edbt14demo,
      author = {Naimdjon Takhirov and Fabien Duchateau and Trond Aalberg and Ingeborg Solvberg},
      title = {KIEV: a Tool for Extracting Semantic Relations from the World Wide Web},
      booktitle = {International Conference on Extending Database Technology (EDBT)},
      year = {2014},
      pages = {632-635},
      publisher = {OpenProceedings.org},
      ee = {http://dx.doi.org/10.5441/002/edbt.2014.61},
    }

  2. An Integrated Approach for Large-Scale Relation Extraction from the Web
    Asia-Pacific Web Conference (APWeb), 2013
    Naimdjon Takhirov and Fabien Duchateau and Trond Aalberg and Ingeborg Solvberg

    @inproceedings {apweb13,
      author = {Naimdjon Takhirov and Fabien Duchateau and Trond Aalberg and Ingeborg Solvberg},
      title = {An Integrated Approach for Large-Scale Relation Extraction from the Web},
      booktitle = {Asia-Pacific Web Conference (APWeb)},
      year = {2013},
      pages = {163-175},
      url = {http://dx.doi.org/10.1007/978-3-642-37401-2_18},
      publisher = {Springer},
      ee = {http://dx.doi.org/10.1007/978-3-642-37401-2_18},
    }

  3. An Evidence-based Verification Approach to Extract Entities for Knowledge Base Population
    International Semantic Web Conference (ISWC), 2012
    Naimdjon Takhirov and Fabien Duchateau and Trond Aalberg

    @inproceedings {iswc12,
      author = {Naimdjon Takhirov and Fabien Duchateau and Trond Aalberg},
      title = {An Evidence-based Verification Approach to Extract Entities for Knowledge Base Population},
      booktitle = {International Semantic Web Conference (ISWC)},
      year = {2012},
      pages = {575-590},
      publisher = {Springer},
      url = {http://dx.doi.org/10.1007/978-3-642-35176-1_36},
    }

Screenshots

KIEV detecting relationship 1
Detecting relationship: extracted patterns
KIEV detecting relationship
Detecting relationship: results
KIEV detect object entity
Detecting object entity
KIEV discovering examples
Discovering exemples (of entities)
workflow of KIEV
Workflow of KIEV