I checked this old question on tools for text corpus annotation and there are many good tools there, however the tools there seem not to cover my problem.

I have a linear text segmentation problem, in which a text needs to be segmented in different sections (topics). I'm building a model for it, but I have no training data. Therefore, the team decided to manually label some text, by annotating blocks in the text that represent each section. I tried some NER or POS labelling tools, but they are not very convenient for selecting several lines and paragraphs to annotate a label.

Is there a good tool for human annotation of text segmentation?

  • $\begingroup$ There is a simple yet powerful tool called BRAT $\endgroup$ – sophros Apr 11 '18 at 10:08

Since I couldn't find a suitable tool, so I made a tool for textual topic segmentation.

To anyone interested in this question, I forked YEDDA's entity recognition project and changed it, accordingly. You can check it here in GitHub: https://github.com/ThiagoSousa/YEEDASeg. My project is called YEDDASeg.

Edit: People mentioned to use BRAT, but it is good to select a word and tag with with a label. My case is a bit more complicated, because it can be many paragraphs. In YEDDASeg, you basic just click on the text and press a key to label a topic in the text. It is simpler for this purpose.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.