Skip to content

natural language processing and computer vision use cases for non technical user

License

Notifications You must be signed in to change notification settings

mmarouen/marabou

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Marabou machine learning toolkit

Natural Language Processing and computer vision API. The goal is to enable non machine learning specialists to leverage the advantages of various nlp/ computer vision use cases using an easy interface. Our goal is to offer a toolkit enabling low cost integration of machine learning capabilities in the business use cases
The API offers 3 scenarios:

  • Sentiment analysis: Deriving sentiments in sentences (positive, negative, neutral), in articles or in a stream of data such as tweets or product reviews.
  • Named entity recognition: Identifies predefined entities in a given text such as Date, Person, Location ...
  • Clothing classification: Identifies clothes type (coat, pants, jean, necklaces ...) from a given input image

Online interface

The application offers 3 use cases each accessibe through an easy to use interface.
Simply go to http://marabou.herokuapp.com/

Sentiment analysis

For a simple interface in which you type an expression and get a classified output based on the text semantics,

  1. Inserting the text you want to analyse
  2. You receive a probability for a positive review

Named entity recognition

You got a text containing several entites you'd like to classify (geographical, data, persons, ...)

  1. Inserting the text you want to analyse
  2. The text will be tokenized and each word will be assigned a category

Fashion classifier

The toolkit support clothing type detection. For that you need to go to "Deep Fashion" classifier. This is a ML program using state of the art deep learning algorithms to detect what kind of clothing is in the image

  1. upload the image you want to detect
  2. the program will analyse the image and feedback the clothin category: jean, skirt, shirt ...

Activate pre-commit checks

If you wish to have your code automatically checked for quality before every commit, you can activate the githook
Under the git repo type git config core.hooksPath .githooks/

Contact and further questions

Please do not hesitate to send your questions/remarks to: [email protected]