Contextual Intelligence Scalability Issues in Personal Semantic Networks

Contextual Intelligence: Scalability Issues in Personal Semantic Networks Oliver Brdiczka

Contextual Intelligence • 85 -95% of digital content is human readable but “unstructured” = untyped, untagged, ambiguous to a computer • Context disambiguates meaning – Context helps interpret words in a sentence, just as it helps interpret information in situations • Content and context analytics convert the ocean of unstructured information, into a Personal Semantic Network

Contextual Intelligence: Harness Content and Context Explicit Inferred from activity • Social tagging • Author • Task, activity • Comments • Relation to other documents • Bookmarks • Relation to people Extracted from content • Topic • Author Content • Document type • People, places, and Topics Tasks People things • Key concepts Events Places • Project • Shared Images Information and people always linked to each other thru context Information self-organizes for faster, everywhere retrieval New Context-Enabled Applications Sensed • Activity/Task • Location • Proximity to other devices • People

Challenges • Silos of Information • Relevance • Latency • Bootstrapping and Longevity • Reasoning Intelligence • Human-Human Interaction / Sharing • Interfacing PARC | 4