2019-06 Graduation assignment MILLION

MILLION: MachIne Learning aLgorithms for It OperatioNs

Figure 1: Looking for this person!

Managing IT operations is a cumbersome task for companies that provide software platforms. System logs play an important role in identifying and resolving issues. Logs are used to track activity of applications and users. If an application faults then error logs are generated that record the cause. System administrators are tasked with taking decisions to resolve issues based on information provided by these errors logs. As the number of entries in these logs increase, it becomes increasingly more difficult for these firms to manually take decisions with regards to all error logs. Hence, companies are interested in applying machine learning in order to facilitate some of these decisions. At the same time, it is important to provide system administrators with creative and adequate visualizations to facilitate decision making.

A company in Enschede is interested in providing you with an opportunity to carry out your Master thesis research with them.


Machine learning algorithms can be used to discover anomalies in these log files. If trained with enough data they can further be useful in categorizing these anomalies. The student has to apply and test the efficiency of various machine learning algorithms for automating error discovery and the rectification process.

For this assignment we are looking for a student with the following qualifications: 

  1. A bachelor degree in BIT (Business and IT), Computer Science, IEM or related field.
  2. Sufficient knowledge of python and/or java.
  3. Some knowledge of NLP and ML algorithms.
  4. Knowledge of system logs, database and elastic search is an advantage.

It is possible for the student to begin working on the assignment asap. In case you are interested in this assignment, please feel free to contact Abhishta (s.abhishta@utwente.nl) or Wouter (w.j.a.vanheeswijk@utwente.nl) .