No description, website, or topics provided.
Scala Python
Switch branches/tags
Nothing to show
Clone or download
Ling-Ling Zhang Ling-Ling Zhang
Ling-Ling Zhang and Ling-Ling Zhang Added evaluation metrics
Latest commit 6502909 Jun 7, 2015

README.md

#CoreNLP Text Classification Engine

This engine uses Stanfords CoreNLP library to classify strings. The engine will place query strings into a class based on the sample data.

##Sample Data

Data files should be formatted as classification::text. Where text can be any arbitrary string (with or without spaces), and classification is a string (without spaces)

$ python data/import_eventserver.php --access_key <your_access_key>

##Queries and Results

Queries must contain a "text" that should be populated using the string you want classified. In addition to this required field you can add the optional fields "gender", "dizziness", "convulsions", "heart_palpitation", "shortness_of_breath", "headaches", "effect_decreased", "allergies_worse", "bad_interaction", "nausea", and "insomnia" to your query where gender takes "male" and "female", and the rest take "yes" and "no". The result contains a single field "queryResults" which will contain your answer in the following format: query ==> classification

Simple query with only text

$ curl -H "Content-Type: application/json" -d '{ "text": "classify me" }' http://localhost:8000/queries.json
$ {"queryResults":"classification ==> class"}

Query with text and some optional arguments

$ curl -H "Content-Type: application/json" -d '{ "text": "classify me", "gender": "female", "heart_palpitation": "no", "bad_interaction": "yes", "headaches": "no" }' http://localhost:8000/queries.json
$ {"queryResults":"classification: class"}