Analogy 0.1

Experimental Open-source Natural Language Processing project for similiarity and difference retrieval

Analogy is an experimental open source project for Natural Language Processing. It aims to perform 2 newly introduced NLP tasks: word comparison and sentence comparison.

Analogy provides semantic similiarity and differences between two pieces of text. Text can be in the form of a word or a sentence.

A pretrained model is released to get started. You can also retrain upon an existing model.

Getting Started:

Prerequisites: Python 3.0 or higher Stanford Core NLP (3.9.2)

Installing:

pip install analogy

Read instructions on how to install and run stanford corenlp server.

Analogy functions:

1. findComparison(model, word1, word2)
2. findSentenceComparison(model, sentence1, sentence2)
3. trainModel(sentences) #Input is list of sentences
4. retrainModel(model, sentences)
5. saveModel(name, model) #Be sure to add '.npz' at last
6. loadModel(name)

Example:

findComparison(model, "apple", "orange")

Output:

Word1 = apple Word2 = orange Similiarity = fruit