I'm going to start a Computer Science phd this year and for that I need a research topic. I am interested in Predictive Analytics in the context of Big Data. I am interested by the area of Education (MOOCs, Online courses...). In that field, what are the unexplored areas that can help me choose a strong topic? Thanks.

  • 1
    $\begingroup$ This is a very broad question. Please try to narrow it down. What are you interested in? $\endgroup$
    – Sean Owen
    Sep 25, 2014 at 21:08
  • 1
    $\begingroup$ Hello Sean, thanks for your answer. I have just edited my question and narrow it down a bit. I specified that I am interested in the area of Education with the new revoluions like MOOCs as an example... $\endgroup$ Sep 27, 2014 at 16:58
  • $\begingroup$ One of the most promising tools for machine learning with big data is MLlib from Apache Spark project. There's already a number of methods implemented, but much more is to be written yet. If you are also interested in education, take some algorithmic problem from it (e.g. predicting student's score or course efficiency) and design solution for that problem. Having told that, I should warn you that for using big data solutions your data should really be pretty large, and I'm not sure MOOCs will give that much data. $\endgroup$
    – ffriend
    Sep 27, 2014 at 22:10
  • $\begingroup$ I would like to thank you first for your detailed answer. Yes you are right, the availability of the so much large data is very crucial. I will try to think about that and get back to you :) Thanks for your proposition of the subject :) $\endgroup$ Sep 28, 2014 at 14:59
  • 1
    $\begingroup$ Our start up is looking for academics to work on big data via stock analysis, if that field interests you, feel free to reach out. $\endgroup$ Mar 3, 2015 at 21:28

2 Answers 2


As a fellow CS Ph.D. defending my dissertation in a Big Data-esque topic this year (I started in 2012), the best piece of material I can give you is in a link.

This is an article written by two Ph.D.s from MIT who have talked about Big Data and MOOCs. Probably, you will find this a good starting point. BTW, along this note, if you really want to come up with a valid topic (that a committee and your adviser will let you propose, research and defend) you need to read LOTS and LOTS of papers. The majority of Ph.D. students make the fatal error of thinking that some 'idea' they have is new, when it's not and has already been done. You'll have to do something truly original to earn your Ph.D. Rather than actually focus on forming an idea right now, you should do a good literature survey and the ideas will 'suggest themselves'. Good luck! It's an exciting time for you.

  • $\begingroup$ Thanks M.Dax a lot ! Yes you are right, thank you very much for the advice. $\endgroup$ Jun 27, 2015 at 15:13
  • Limitations of ML-Powered Predictive Models

  • Predictive Models on Conversion Studies for the Level Premium Term Plans

  • Data Management

  • Business intelligence(BI) predictive Analytics

  • Big data and Market Intelligence

  • Educational data mining

  • Automation of predictive models

  • predictive data quality using ML

  • Open source Data Mining

  • Data Mining System

  • Data mining Trends

  • Health data mining

  • Web Analytics solution

  • Data Mining marketing

  • Data Mining in Search Engine Analytics

  • Fraud & Security

  • Risk Management


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.