The corpus contains 759 sentences, 2268 Named Entities, 15213 words and 24839 characters.

The Named Entities are labelled in the "BIO" format.

All the data are in the Construction-NER-corpus.csv, where words are listed in the first column, while tags in the sencond.

The raw corpora is collected from a series of supervison documents of a contruction project.

The principles guiding the annotation are in Specification for annotation.docx or Specification(in English) for English version.

Kappa test were applied on the corpus, below is the result:

      B     I     O
  B   2026  119   123
  I   70    2763  174
  O   18    46    9874

If you have any question, please feel free to contact us: qqz@zju.edu.cn