subhashreehazra / exam_auto

exam_automation

The goal is to automate the paper correction process even for lengthy answers for different subjects.

THE END GOAL

Building a conceptual , semantic similarity machine learning model and wrap a scoring above it. Plug it wih the web-app and develop a strategy for model maintenance and monitoring The current experiment is being done for Operating Systems subject in SRM university

MODEL CONSTRUCTION USING SIAMESE-LSTM

Inspired from Tensorflow Implementation of https://github.com/dhwajraj/deep-siamese-text-similarity

Referrences:

https://colah.github.io/posts/2015-08-Understanding-LSTMs/
https://colah.github.io/posts/2014-07-NLP-RNNs-Representations
Using MaLSTM model(Siamese networks + LSTM with Manhattan distance) to detect semantic similarity between question pairs. Training dataset used is a subset of the original Quora Question Pairs Dataset(~363K pairs used).
FOR THE SAKE OF MODEL CONSTRUCTION AND SELECTION, WE ARE USING KAGGLE'S QUORA QUESTION PAIRS TO TRAIN THE MODEL INITIALLY, WHILE THE DATA COLLECTION IS BEING DONE.
Data
GoogleNews-vectors-negative300.bin.gz
Kaggle's Quora Question Pairs Dataset
Kaggle's test.csv is too big, so I had extracted only the top 20 questions and created a file called test-20.csv and It is used in the predict.py.

DATA COLLECTION

OVERVIEW :

This is a common platform for simulating and uploading sentence pairs and answer keys for model training and testing.

Features:

Add Subject and exam: Feature to add any subject and exam
Add Question: Feature to add question for given subject for exam/simulation
Simulate Question: Fetch the questions with type = simulation and simulate different sentences for the answers with various type of variations and upload back to the training collection.
Type Answers: For given question with type = exam upload the answers written by students for that question
Correct Answers: For a given student, given exam, given question, correct answers. Match the following sentences pairs and submit back to test collection set.
LINK TO THE PORTAL: http://157.245.100.55/

Data collection has already been done by collecting examination answers written by students.

WEB PORTAL OF THE APPLICATION