Abstractive Text Summarization

This repository aims at providing Stacked LSTM-based Encoder-Decoder implementation for Abstactive Text Summarization task, which generates new relevant sentence as summary for each input (original text in the form of longer paragraphs). This task has many applications:-

• Entity timelines - Given an entity and a time period, provide a summary of the most memorable events involving this entity.
• Sentence Compression - Given a sentence, generate a shorter one while preserving the essential contextual information.
• Summarization of User Generated Content - Identify main aspects and summarize for machine or user consumption.

Requirements

• Tensorflow
• keras
• NLTK
• NumPy
• Pandas
• Matplotlib

Usage

Data

• The developed model is trained on the Amazon Fine Food reviews dataset, which can be downloaded from here.
• The data spans a period of more than 10 years, including all ~500,000 reviews up to October 2012.
• These reviews include product and user information, ratings, plain text review, and summary.
• It also includes reviews from all other Amazon categories.

Training and Testing

To train the model and generate summary for each new test paragraph, run Abstractive_Text_Summarization.py.

Results

Following are a few summaries generated by the model:-

Review: aware decaf coffee although showed search decaf cups intended purchase gift kept recipient drink caffeine favorite means 
Original summary: not decaf 
Predicted summary:  decaf decaf

Review: first time using fondarific fondant general one really easy use baby shower cake worked indicated also colored made two tier cake final product looked great greasy 
Original summary: easy to use 
Predicted summary:  great product

Review: size quite good dog training smell strong cannot put open bag must seal everytime gave treat otherwise dog stand trying fetch believe taste great puppy purchase sure 
Original summary: strong smell and my puppy loves it 
Predicted summary:  my dog loves these

Review: daughter drinking since months old months old still loves snack time healthy delicious great addition menu 
Original summary: great snack 
Predicted summary:  great snack

Review: love stuff great store bought homemade baked goods kicking things professional level works colored dark light frosting also used dusting powdered sugar pretty fine texture 
Original summary: fun like dust 
Predicted summary:  perfect