Hello, how to generate the following file？

[zhangsanaixuexi]

q.txt: The file of questions.
te.txt: The file of topic entities (Obtained via NER tool for named entity detection and Google Knowledge Graph Search API).
con.txt: The file of detected constraints, which usually cannot be detected as topic entities (Obtained via a simple dictionary constructed on training data).
a.txt: The file of answers.
g.txt: The file of ground truth query graphs (No such file is provided for ComplexQuestions).

[lanyunshi]

Hi,

These are files that we generated via our pre-processing step, either extracted from the raw dataset file or generated via some simple steps (such as te.txt, con.txt). For g.txt file, we obtained it by selecting the relation path with the maximum F1 score, but it's not very important as we didn't use g.txt file in our code.

[Sarahtorekryo]

Hi,

These are files that we generated via our pre-processing step, either extracted from the raw dataset file or generated via some simple steps (such as te.txt, con.txt). For g.txt file, we obtained it by selecting the relation path with the maximum F1 score, but it's not very important as we didn't use g.txt file in our code.

Hi, thanks for your extraordinary work! Could you please tell what entity linking tool did you use to get the ranked topic entities and their scores in the question(contents in te.txt)? Thanks a lot!!

[imjinshuo]

Sorry, I still do not know how to generate the con.txt after reading your answering.