ParallelDots Ruby API

A wrapper for the ParallelDots API.

Installation

From Gem:

gem install paralleldots

API Keys & Setup

Signup and get your free API key from ParallelDots. You will receive a mail containing the API key at the registered email id.

Configuration:

> require 'paralleldots'

# Setting your API key
> set_api_key("YOUR API KEY")

# Viewing your API key
> get_api_key()

Supported APIs:

• Semantic Similarity
• Sentiment Analysis
• Taxonomy
• Named Entity Extraction/Recognition ( NER )
• Keywords
• Intent
• Emotion
• Abuse
• Multiple Language Sentiment
  • Portuguese ( pt )
  • French ( fr )
  • Spanish ( sp )

Examples

> require 'paralleldots'

> similarity( "Sachin is the greatest batsman", "Tendulkar is the finest cricketer" )
{"actual_score": 0.8429316099720955, "normalized_score": 4.931468684177398, "similarity": 4.931468684177398}

> sentiment( "Come on, lets play together" )
{"sentiment": 0.8513014912605286}

> taxonomy( "Narendra Modi is the prime minister of India" )
{"tags": [[u"finance", 4.088], [u"government", 3.4284], [u"business", 1.2719]]}

> ner( "Narendra Modi is the prime minister of India" )
{"entities": [[u"Modi", 1.0, [u"person"], u""], [u"India", 1.0, [u"org"], u""], [u"Narendra", 1.0, [u"org"], u""]]}

> keywords( "Prime Minister Narendra Modi tweeted a link to the speech Human Resource Development Minister Smriti Irani made in the Lok Sabha during the debate on the ongoing JNU row and the suicide of Dalit scholar Rohith Vemula at the Hyderabad Central University." )
{"keywords": [[u"Human Resource Development Minister Smriti Irani", 6], [u"Prime Minister Narendra Modi", 4], [u"Hyderabad Central University", 3], [u"ongoing JNU row", 3], [u"Dalit scholar", 2], [u"Lok Sabha", 2], [u"Rohith Vemula", 2]]}

> emotion("Did you hear the latest Porcupine Tree song ? It's rocking !")
{"emotion": "happy"}

> intent("Finance ministry calls banks to discuss new facility to drain cash")
{"intent": "news"}

> abuse("you f**king a$$hole")
{"confidence_score"=>0.998047, "sentence_type"=>"Abusive"}

> multilang_sentiment("La ville de Paris est très belle", "fr")
{"sentiment": "positive", "confidence_score": 0.998047}