• 

  MLH Quizzet

• 

  Quiz generated from uploaded text

• 

  Results Page

MLH Quizzet

This is a smart Quiz Generator that generates a dynamic quiz from any uploaded text/PDF document using NLP. This can be used for self-analysis, question paper generation, and evaluation, thus reducing human effort.

Features

• implements automatic question generation (AQG) techniques > Automatic question generation (AQG) is concerned with the construction of algorithms for producing questions from knowledge sources, which can be either structured (e.g. knowledge bases (KBs) or unstructured (e.g. text))
• helps in resource saving(time, money, and human effort)
• enables the enrichment of the teaching process, adapt learning to student knowledge and needs, as well as drill and practice exercises
• presents an automatic mechanism to assemble exams or to adaptively select questions from a question bank

WorkFlow

Input

• Input in the form of a text/PDF file that consists of English text data
• The English text must preferably be over a single broad topic with multiple smaller subtopics
• The helps in generating a good quiz for the user to practice on

Text Pre-processing

• Text is pre-processed so it can be in a format as expected by the natural-language models
• non-alphanumeric characters(except full stops) are dropped
• This also helps improve the output of the natural language model

Named Entity Recognition + Entity Ranking

• Spacy’s NER model is used to find the named entities from the given text. These consist of people’s names, dates, places, quantities, etc.
• These entities are good candidate questions and are ranked based on their TF-IDF score ( a metric used to weigh a word across multiple documents )

Incorrect Option Generation

• A Word2Vec model implemented in gensim is used to find the top 10 similar entities for a given entity. We then pick the least 4 entities as alternate options.
• We can also pick words from the given text itself if the entity is not present in the model vocabulary

Technology Stack:

• Frontend: HTML, CSS, Vanilla JS
• Backend: Flask
• IDE: VS Code
• Design: Canva
• Version Control: Git and GitHub
• Database: Sqllite3

Browser Support

• Firefox: version 4 and up
• Chrome: any version
• Safari: version 5.2 and up
• Internet Explorer/Edge: version 8 and up
• Opera: version 9 and up > Note: Support for modern mobile browsers is experimental. The website is not responsive in mobile devices until now.

MLH Fellowship( Fall 2020)

This is a hackathon project made by MLH Fellows(Fall 2020) - Pod 1.0.0 i.e. Fantastic Falcons

Team:

"Alone we can do so little; together we can do so much."

S.No.	Name	Role	GitHub Username:octocat:
1.	Pragati Verma	Frontend Developer	@PragatiVerma18
2.	Kshitij Kotasthane	Backend Developer	@kshitij86
3.	Vignesh S	ML	@telescopic

Contributors ✨

Thanks goes to these wonderful people (emoji key):

Pragati Verma 💻

Kshitij Kotasthane 💻

Vignesh S 💻

This project follows the all-contributors specification. Contributions of any kind welcome!

---

Built With

• complexityintelligence-named-entity-recognition
• css
• flask
• html
• javascript
• jinja
• mlh
• natural-language-processing
• python
• quiz-generator
• spacy
• werkzeug