Skip to main content

An off-the-rack NLP sentiment classifier- upload your own corpus or use the pre-installed ones

Project description

# empathyMachines
> A standalone NLP sentiment classifier you can import as a module

## Purposes

1. Offer a batteries-included NLP classifier you can use either on it's own, or to make sentiment predictions as part of a broder NLP project (for example, when classifying customer messages, whether the customer is angry or not might help you determine if this is a compensation request, or a request to adjust their address.)
1. Have the entire sentiment prediction process scaffolded so you can feed in your own training corpus, and easily train an NLP sentiment classifier.

## How to use

1. `pip install empythy`
from empythy import EmpathyMachines
nlp_classifier = EmpathyMachines()

### Corpora included

#### NLTK Movie Reviews
The classic sentiment corpus, 2000 movie reviews already gathered by NLTK.

#### Assembling a custom Twitter sentiment corpus
[CrowdFlower]( hosts a number of Twitter corpora that have already been graded for sentiment by panels of humans.

I aggregated together 6 of their corpora into a single, aggregated and cleaned corpus, with consistent scoring labels across the entire corpus. The cleaned corpus contains over 45,000 documents, with positive, negative, and neutral sentiments.

### Train on your own corpus

Feel free to train a classifier on your own corpus!

Two ways to do this

1. Read in a .csv file with header row containing "sentiment", "text", and optionally, "confidence"
- Pass the name of the .csv file to train, like so:
- `nlp_classifier.train(corpus='custom', corpus_path='path/to/custom/corpus.csv')`
1. Pass in an array of Python dictionaries, where each dictionary has attributes for "sentiment", "text", and optionally, "confidence"
- `nlp_classifier.train(corpus='custom', corpus_array=my_array_of_texts)`
- Two important parts to this, both `corpus='custom'`, and `corpus_array=my_variable_holding_the_documents`.

### Advanced Usage
1. `nlp_classifier.train(verbose=False)` to turn off print status statements while training.
1. `nlp_classifier.train(print_analytics_results=True)` to print out results of training the classifier.

Project details

Release history Release notifications

This version
History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
empythy-1.0.0-py2.py3-none-any.whl (1.6 MB) Copy SHA256 hash SHA256 Wheel py2.py3 Aug 29, 2016
empythy-1.0.0.tar.gz (1.6 MB) Copy SHA256 hash SHA256 Source None Aug 29, 2016

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging CloudAMQP CloudAMQP RabbitMQ AWS AWS Cloud computing Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page