Coursera Personalized Recommendations par Abhinav Agarwal

Coursera Ontology-Based Personalized recommendation system is a research project being undertaken by Abhinav Agarwal and Divyansh Mishra, 2 students of ICT Department at MIT Manipal, under the supervision of Dr. Sucheta Kolekar.

The course completion rates on the Coursera platform are currently less than 10% of all the enrolled students. We are working to provide targeted and personalized recommendations to Coursera users, such as how difficult the upcoming quiz is going to be? Or how many revisits learners like you made to a particular video? Existing systems only partially provide such recommendations. We are working to build a new recommendation system, which would aid students in learning more from the online courses.
As you can already guess, for achieving this goal, we need some data regarding the way you and other users learn from the Coursera platform.
This extension has been developed to facilitate willing students and other users of Coursera to share their data with the above authors for the speedy progress of the research project. Please find below some of the FAQs regarding data collection that you may have. For any more questions, always feel free to drop an mail to the authors.



This extension is primarily intended for students at MIT Manipal.

Q) What information will you collect?
A) This extension will not collect any personally identifiable information. We only collect SHA 256 hashes of your name and user id (making you totally anonymous for us). Other than that, various content usage data of Coursera, such as the videos watched or quiz scores or participation in discussion form, etc. will be recorded.

Q) What do I need to start sharing my data?
A) Just install the extension and forget :). The extension will collect all the information unobtrusively, on its own. The extension primarily scrapes the Coursera website for your data. By installing this extension, you are allowing us to collect your data.

Q) Will this extension impact my browsing experience in any way?
A) No. This is a very lightweight application, which will work well, even on a very resource-constrained machine. The app doesn't contain any intrusive pop-ups or ads or something similar.

Q) Will you collect any data other than mentioned above?
A) No, the data collection is strictly restricted to Coursera only, as described above.

Q) How will you keep my data safe? Will you sell my data to someone else?
A) The backend of this application is entirely hosted on servers managed by Google, which are among the safest places where your data could be. There is always end-to-end encryption during transmission to uphold the highest levels of privacy.
We are collecting data strictly for the Coursera Ontology-Based Recommendation System, and we don't have any commercial intentions, so the question of selling your data doesn't arise at all.

Q) Do you have permission to collect this data?
A) Primarily, we are authorized by the college to collect this data. But more importantly, you are the owner of your data, and we only need your permission to collect your data.

Q) When will the research project be over, and what are the expected results?
A) While there is no specific deadline, we are aiming for a completion date of April 2020. The final product will give never seen before usage-based personalized in-course recommendations for the Coursera platform.



Right now, this extension only collects user data. But, as time progress, we will slowly start rolling out recommendations through this extension.


If you have any other questions, please feel free to drop an mail to the authors.