Deliverable D4.1 “Development of Automated Techniques to Detect Early Indications of Malicious Behavior of Social Network Users”

One of the main objectives of the ENCASE project is to develop a browser add-on and its corresponding machine learning algorithms for the detection of malicious and problematic behavior (such as cyberbullying, sexual predators, distressed or aggressive behavior, etc.).

During this project, we attempt to identify and quantify the various types of online abuse and we research a number of methodologies to detect hateful content, raid, abuse and bullying. Furthermore, to train our algorithms, we are also collecting a large-scale crowdsourced dataset.

This deliverable describes in detail the significant progress towards automatically detecting malicious behavior in the context of tasks T4.1: “User profiling to detect and prevent malicious and criminal activities” and T4.2: “Sentiment and affective analysis on individual and collective basis”.

The purpose of T4.1 is to collect and analyse social network data, user profiles and other relevant data aiming to understand how users are exposed to predators, cyberbullying and other types of malicious behavior. Using the extracted features, we build classifiers that will be used later by the corresponding browser add-on to detect multiple types of malicious behavior.

In T4.2 we perform sentiment and affective analysis for more in-depth understanding of user behavior. From the acquired knowledge, we devise feature extraction techniques and we build classifiers to early indications of malicious behavior of social network users.