Named Entity Recognition - direct matching with a dictionary

user_9498549 0
Viewed 999

I would like to use named entity recognition (NER) to identify words or phrases in the text which align with clinical concepts.

I have a dictionary that contains the description of a diagnosis and its label code. Example of first 2 rows:

ICD10  ICD10Term
----------------
A00    Cholera
A000   Cholera due to Vibrio cholerae 01, biovar cholerae

Firstly, I wanted to start with a direct match with the dictionary but I am not sure how I should do it. Should I just search if the dictionary contains a certain string? Or should I use a package or a tool?

I found spacy-lookup : https://github.com/mpuig/spacy-lookup

I am not sure how I would use this package at a large scale as I have many txt files and the dictionary.

Also, is there any other package\tool that you would recommend me?

named-entity-recognition spacy python
2 Answers

You'll have to first train your own NER model to do that.

Prepare your dataset as spaCy requires and then train your model. Then your model should be able to detect the entities.

Admin1
user_32613010
Related