Triples, Named Entity Recognition and All that Jazz

“Okay, Knowledge navigator (wink, wink, to Apple’s concept from 1987),” I imagined saying while I preparing this edition of the newsletter about Content and Knowledge Graphs, “play some Chet Baker, and then give me a musician he was influenced by and also possibly played with.”

“No problem!”, said the system, as it was able to acces the API Linked Jazz, retrieve a Youtube video with Chet Baker and further give me information about Sam Rivers from an interview with him that was transcribed and codified with semantic data.

Now, exit my imagination, and enter the real, tangible world of Linked Jazz project – yet another graph of knowledge built with semantic technologies.

Linked Jazz Project

Linked Jazz is a project founded in 2011 under the umbrella of the Semantic Lab at Pratt and aimed at presenting visually a network of jazz musicians, instruments, genres and works through Linked data.

Technically, Linked Jazz uses linked open data (LOD) to support the visualization of a digital network where people can explore personal and professional relationships of musicians. Utilizing documents from digital archives of jazz history, Linked Jazz works to expose the social relationships between jazz musicians and reveal their community’s network.

Recently Linked Jazz added the transcripts of interviews with musicians in which more relationships are being revealed.

The Linked Jazz Oral History Network was one of the first outcomes of the project. It is a dataset and a network visualization based on 54 transcribed oral histories–interviews with musicians and other figures from the jazz world. The project members are currently working with jazz archives from Carnegie Hall and Tulane University to integrate and enrich our RDF dataset with new semantic layers Semantic Lab (see “Projects.” Linked Jazz, n.d. https://semlab.io/projects/#linked-jazz).

It Don’t Mean A Thing If It Ain’t Got That Semantics

Entering the vizualization tool https://linkedjazz.org/network/, a user can start a journey through a richly interlinked universe of jazz players. The facts are connected via meaningful relationships, who knew who, who played with whom, who was influenced by whom.

A user can browse YouTube videos, featuring interviews, reveal connections between musicians and further dive into unknown links between artists, works and songs.  Literally straddling nodes, node after node, anyone interested in deepening their understanding of a given topic or a given jazz musician, can follow typed links to reveal and explore connections.

The interconnected Linked Data empowers in turn empowers the use to uncover meaningful connections between documents and data related to the lives of jazz musicians and to their work and the influences they had in the rich and diverse social networks they eixsted.

Not only that buu the project has built a way for people to add the wisdom of the crowd.

Linked Jazz 52nd ST is a platform where users can select the type of relationship shared by jazz musicians based on short excerpts of interview transcripts pulled from jazz history archives. The relationships that users choose contribute to the Linked Open Data semantics on these jazz musicians.

Epilogue and a vision for linked marketing communications

By codifying (existing and crowdsourced) knowledge about musicians’ connections—such as mentorships, collaborations, and encounters— Linked Jazz created a structured and dynamic dataset that computers can analyze. This in turn allows end users – be they researchers, historians, or jazz enthusiasts to explore the social fabric of jazz, identify influential figures, and better understand how musical ideas and careers have evolved over time. Not only that but also by transforming these relationship into Linked Open Data, the platform created a corpus of sorts that can in the future make these connection accessible and usable across different digital platforms.

If we imagine such connectivity at the level of marketing content, that would be a platform where users could navigate through content pathways of their own choice, following their own pace and enjoying the benefit of pulling content, topics and connections that matter to them on their own terms.

For example, imagine the content from all the webinars you have produced as a brand linked in the way Linked Jazz is linked. With your speakers, their topics, the mentions between the various webinars and the products and services presented in them.

What a wonderfil world!

Let’s work and link data towards building more of it!

Thanks for reading this story, which is an unplugged version of the Knowledge Graphs stories I present in my book Being Dialogic. I would love to hear your thoughts on the subject of webby words.