Abstract:
The explosion of modeling complex systems using attributed networks boosts anomaly detection research in such networks, which can be applied in many high-impact domains. Many existing attempts simply concatenate multiple views into a feature vector, ignoring the statistical incompatibility between heterogeneous views. Multi-view data can help detect anomalies better than single-view data. The abnormal patterns naturally behave differently in different views, which matches people’s desire to discover specific abnormality according to their preferences for views (attributes). Most methods don’t account for user preferences, so they can’t adapt. Thus, we propose a multi-view framework Alarm to incorporate user preferences into anomaly detection and simultaneously address heterogeneous attribute characteristics through multiple graph encoders and a well-designed aggregator that supports self-learning and user-guided learning. Disney, Books, and Enron datasets show Alarm’s improvement in detection accuracy as measured by the AUC metric and its effectiveness in supporting user-oriented anomaly detection.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
