Learning analytics is a new discipline which aims at using statistical, econometric and machine learning methods to analyze a great amount of data about students’ and schools’ activities and results. The ability of elaborating and interpreting this data could activate change within schools/universities, with the aim of establishing practices and interventions which can improve educational results.

Evidence reveals that educational managers are still not making the most with the potential of data they have at their disposal. Moreover, there is still room to analyze of data can influence decision-making processes within school/universities and how they can be leveraged for assessing the effectiveness of specific interventions and practices. Another important direction for Learning Analytics is the development of predictive models, that is to say statistical and machine learning models which are able to predict students’ performance in advance (i.e. identifying most promising students or those who are at-risk to not succeed).

In the thesis, the student(s) will conduct the following activities:

• A systematic review of the academic literature which defines the models and methods used in Learning Analytics
• A study of practical experiences of learning analytics, in other words they will conduct case studies about specific schools/universities which developed initiative of LA in past years
• An experiment of using LA techniques and methods using data from a real-world experience (for example, using data from Politecnico di Milano)