Project: Plagiarism checker

Mentor: Soumendra Kumar Sahoo

Problem Statement

Develop a plagiarism checker for English language.

Requirements

• There should be an API service in which a user can paste a paragraph or document.
• The service will search for the matching sentences/document found across the internet and if there is a close enough match(es) found, the service will provide:
  • the matched sentences/paragraphs of the pasted text,
  • the copyright of the original text and
  • the source link with the original date of publication where it matched.

How to proceed

Source documents Procurement

• Set the scope of the project. What is the internet for you to search for?
• The source should have the above mentioned requirements, which you should be able to provide in result.
• To extract data from websites you can check this Scrapy tutorial explained in Odia.

Storing the data

• Find out the ways you can store the text with corresponding metadata.
  • DB or file? SQL or NoSQL; why?
• Find out how can you store the source text such that, you can easily match the query text by user with the large number of source documents. This is they key to the project.
  • Text preprocessing
  • Word to word match? Sentence to sentence match? paragraph to paragraph match? What else?

Create a service

• Create a service to let the user paste the query text
• Search the source documents for the query text and return the results.

Bonus points will be given for

• If you can handle multiple varieties of query text and still detect plagiarism.
• If your service can provide search results in less than 2 seconds.
• If you able to deploy the service for others to try.
• Can you reuse your project to create a plagiarism checker for Odia?

---

Last update: 2022-11-04