SM3: Social and Multimedia Metadata Manager

The SM3 platform provides a multimedia metadata annotation and cross-reference capability designed to front a DAMS/VMS system.  The system provides content enhancement through metadata extraction, enhancement, and cross-referencing using NLP, Search, and Ontology and concept extraction approaches.

SM3 Social and Multimedia Metadata Manager pdf16 SM3 Presentation pdf16
   
SM3-arch

SM3 is provided as a core with domain-specific modules supporting enterprise-level content management, social network (blogs, wikis etc.) integration with user-generated content, email, RSS newsfeeds, and other forms of structured and unstructured text.

The SM3 platform core provides a document analysis, multimedia annotation and cross-referencing capability; including deep-linking into multimedia frame and scene content and annotation with Geo- and time-stamping of frames, scenes and related entities. SM3 supports multiple media types and interfaces, including MPEG7/21.

The metadata extraction, annotation and cross-referencing using NLP, Search, and Ontology and concept extraction approaches enhance content value by identifying and exposing domain entities (persons, organisations, relationships, activity types etc.) for search indexing and analysis purposes.

SM3 provides a module to support intelligence analysis of documents and ELINT traffic with situational awareness support to link user queries to the appropriate multimedia and document content.  This module provides concept analysis and supports the user-interface capability of the  SemanticFusion platform.

For example, SM3 could cross-reference sensor logs, UAV video and audio commentaries, permitting queries by keyword, time, location etc.

Other modules can integrate with automated video and audio analysis systems, applying NLP and concept analysis cross-referencing against other sources to extend and improve annotations, improving search accuracy.

SM3 is currently in development.