Time-varying Knowledge Graphs and Datasets

Semantic Web Interest Group,

Hello. I am excited to share a note with some ideas about time-varying knowledge graphs and datasets.

https://github.com/AdamSobieski/Narratology/blob/main/Content/Time-varying%20Knowledge%20Graphs%20and%20Datasets.md

The note explores ways that RDF Turtle and TriG (without any blank lines) can, today, be used in WebVTT cues. Towards even more efficient storage and transmission, the note also explores syntax possibilities for WebVTT to be able to express prefix declaration blocks for use across multiple cues.

With respect to using linked data (JSON-LD) in cues' payloads, I found a previous work from 2014: Weaving the Web (VTT) of Data [1].

A novel contribution in the note shared, today, involves that time-varying knowledge graphs and datasets can be assembled by merging the payloads of media resources' metadata tracks' instantaneously active cues. A media resource, then, can have, per instant, a knowledge graph or dataset formed from merging those instantaneously active cues' payloads in a metadata track. The note explores ways to assemble these resultant graphs or datasets efficiently, during playback, as cues are entered and exited, using multiset data structures.

Considered use cases include: education, accessibility, artificial intelligence, computational narratology, and animated data visualization and infographics.

Hopefully interesting! Thank you for any comments, feedback, questions, or ideas.


Best regards,
Adam Sobieski

[1] Steiner, Thomas, Hannes Mühleisen, Ruben Verborgh, Pierre-Antoine Champin, Benoit Encelle, and Yannick Prié. "Weaving the Web (VTT) of Data." In 7th Workshop on Linked Data on the Web. 2014.