Engineering is the project coordinator of a pool of 18 partners (companies, institutes, universities, ...) from 10 countries.
The framework will consist of tools to automatically manipulate, analyse and fuse vast amounts of heterogeneous data acquired from different sources including CCTV surveillance content, confiscated desktops and hard disks, mobile devices, Internet, social networks, handwritten and calligraphic documents.
The type of data considered includes text, images, video, audio and biometric information in multiple formats. In order to manage the results of the automated processing, a knowledge repository will be built. It will consider explicit analyst-knowledge and critical legacy information from previous cases. The proposed knowledge representation framework will also allow the system to provide recommendations to analysts, guide the investigation process and perform inference based on evidence extracted from available data.
In LASIE, search and retrieval of evidence will be enhanced through the provision of complex query formulations and multimodal search mechanisms yet through a user-friendly, user-centric human-computer interface.
The aim is to link and merge heterogeneous data retrieved from multiple sources to improve the knowledge-base and the accuracy of retrieved results. The envisaged user-friendly interface will allow analysts to easily visualise and navigate through the retrieved evidence, highlight relevant events and provide feedback to refine their search criteria.
LASIE will follow a privacy-by-design approach, ensuring that all the aforementioned functionalities and the used forensic data strictly obey all legal and ethical restrictions and national laws. This ensures that the outcomes of the system will be accepted in European courts of law.