This repository houses the code for my 3rd year dissertation project. My project idea was to develop an exam revision tool that helps students to work effectively with past exam papers. The software tool is designed to extract useful information and content from past papers and subsequently present this to the user in a standardised and interactive manner. Specifically, the tool includes functionality such as looking up an exam’s rubric details in a standardised format, viewing a past paper’s questions individually, and seeing similar questions to a chosen question.
At a technical level, fulfilling this functionality requires parsing past paper files in PDF format to extract text and identify questions, together with the use of information extraction, natural language processing and machine learning techniques to process the extracted data. Meanwhile, the front-end user interface serves the data via a web application, incorporating the use of web development technologies in addition to interface and interaction design.
The software tool was developed to work with the corpus of available past papers for modules within the Computer Science department (‘DCS’) at Warwick. In this way, the project was intended to yield a proof-of-concept to demonstrate what can be achieved in the extraction of data from past paper materials and how this could be applied in a wider exam revision context such as at GCSE or A-level.
A demo of the web app is hosted on Heroku and is available to view here: https://pastpaperpal.herokuapp.com/. My final report for the project is also available to view here: https://goo.gl/yfnkJx.