search

aiforsec / CyNER

Cyber Security concepts extracted from unstructured threat intelligence reports using Named Entity Recognition
MIT License
79 stars 28 forks source link

Model not workling as intendet #6

Open l0renor opened 1 year ago

l0renor commented 1 year ago

For me, the model isn't working as intended.

For the input

text2 = "Proofpoint report mentions that the German-language messages were turned off once the UK messages were established, indicating a conscious effort to spread FluBot 446833e3f8b04d4c3c2d2288e456328266524e396adbfeba3769d00727481e80 in Android phones."

I get the following output:

Mention: Proofpoint report mentions that the German-language messages were turned off once the UK messages were established, indicating a conscious effort to spread FluBot 446833e3f8b04d4c3c2d2288e456328266524e396adbfeba3769d00727481e80 in Android phones., Class: LABEL_1, Start: 0, End: 246, Confidence: 0.62

Problem 1: The whole text is labeled as one entity. For bigger text, multiple big sections are detected.

Problem 2: Not the original labels but Label_1 is returned.

annous28 commented 1 year ago

Hi,

I am facing the same issue. The model is not extracting named entities.

phretor commented 1 year ago

Please check this comment: https://github.com/aiforsec/CyNER/issues/2#issuecomment-1148494177