NTCIR Math Information Retrieval

NTCIR 10, 11 and 12 had tasks concerning mathematical information retrieval. These tasks primarily focused on keyword search. The corpus was based on arXiv/arXMLiv and Wikipedia. A more detailed description of the NTCIR mathematical information retrieval tasks and the lessons learnt was published at https://doi.org/10.1007/978-981-15-5554-1_12.

Access and Download

Like for the arXMLiv corpus, access is currently restricted to SIGMathLing members but everyone is welcome to join (after signing the SIGMathLing Non-Disclosure-Agreement).

If you are a member, you should have access to the two repositories containing the data:

Citing this resource

  title = {NTCIR math information retrieval task data},
  url = {https://sigmathling.kwarc.info/resources/ntcir/},
  note = {SIGMathLing -- Special Interest Group on Math Linguistics},
  year = {2021}