wikionto

Wikipedia-based ontology engineering

Projects

This project addresses multiple challenges:

• How to recognize relevant articles based on a learned decision tree classifier?
• How to recognize subclasses based on clustering?
• How to recover classification information only by processing natural language text?

Discovering Indicators for Classifying Wikipedia Articles in a Domain

This information is related to the paper accepted at SEKE2019 with regards to the challenge How to recognize relevant articles based on a learned decision tree classifier?

@inproceedings{HeinzLA19,
  author    = {Marcel Heinz and
               Ralf L{\"{a}}mmel and
               Mathieu Acher},
  editor    = {Angelo Perkusich},
  title     = {Discovering Indicators for Classifying Wikipedia Articles in a Domain
               - {A} Case Study on Software Languages},
  booktitle = {The 31st International Conference on Software Engineering and Knowledge
               Engineering, {SEKE} 2019, Hotel Tivoli, Lisbon, Portugal, July 10-12,
               2019},
  pages     = {541--706},
  publisher = {{KSI} Research Inc. and Knowledge Systems Institute Graduate School},
  year      = {2019}
}

• We moved datasets and original scripts to the branch 'seke19'.
• Current datasets and evaluation data can be found at https://github.com/softlang/wikionto/tree/seke19/data/datasets.
• Added natural language tests based on Stanford Parser. CoreNLPServer must be started before using the tests.
• See https://github.com/softlang/wikionto/tree/seke19/README.md for more instructions and manuals for reproduction.