Assuming it is for UG, you may do something in text processing
Example: Sentiment Analysis, Spam Detection in Twitter.
There are many problems from various domains, to give you an example from academics you might want to find which students will drop out. Many tasks also focus on image data.
Go through this for some real problems http://kaggle.com/
I am writing to inform you that I read an article about your topic.
It is article by Haldun Aytug , Serpil Sayın (2009) "Using support vector machines to learn the efficient set in multiple objective discrete optimization" European Journal of Operational Research 193 pp. 510–519
"Theoretical works on the use of digital devices for learning have focused on the affordances users perceive in these devices. Design research has also shown us that learning environments enriched by digital technologies are extremely complex and should be viewed as learning ecologies."
From:
Steffens, K., Bannan, B., Dalgarno, B., Bartolomé, A. R., Esteve-González, V., & Cela-Ranilla, J. M. (2015). Recent Developments in Technology- Enhanced Learning: A Critical Assessment. RUSC. Universities and Knowledge Society Journal, 12(2). pp. 73-86. doi:http://dx.doi.org/10.7238/rusc.v12i2.2453