Method and system for anomaly detection using multimodal knowledge graph

Abstract

A method and system for detecting anomalies associated with a plurality of data objects of a technical installation is provided. The method includes receiving a request to detect anomaly associated with the plurality of data objects in an image data. The image data can be a static image data or a dynamic image data. The method includes classifying the plurality of data objects using a multimodal knowledge graph. The multimodal knowledge graph includes a plurality of pre-classified data objects, each of pre-classified data object is connected using one or more paths. Further, the method includes detecting whether anomaly exists in the plurality of classified data objects. The anomaly is detected by evaluating the plurality of classified data objects with reference data corresponding to the plurality of classified data objects obtained from the multimodal knowledge graph. Furthermore, the method includes generating a notification indicating the detected anomaly based on the evaluation.