International Salmon Data Laboratory - Scott Akenhead

(Scott Akenhead) #1

Greetings from the Pacific Biological Station on Vancouver Island, BC, Canada! I announced the ISDL at GraphConnect 2018-09-20 (no link to video yet). Great reception. Amazing response!
We have (today, more soon) ~20 experts re neo4j tools and "data mobilization" and ~20 domain experts with data on salmon who want salmon "information flow" to leapfrog to 2022 technology via irresistible examples. From field data (new and historical) to decision support products. The catch is sophisticated tools that are sufficiently easy to use and rewarding for unsophisticated users. The revolution begins with salmon but the examples and solutions will be widely applicable.
Your work can impact this, and we will be careful about acknowledgments. First, we need a technology strategy. Then DB design. Then examples of data, tools, and workflows. Then more participants, more data, more examples... rinse, repeat!

(Tom) #2

Welcome to the forum Scott !

tries to resist ... fails ... salmon is going to leapfrog ... absolutely, I'm all in :slight_smile: !


(Michael Hunger) #3

Welcome Scott and I hope the Graphs4Good program is helping you achieve your goals. Let us know if you need help.

(Scott Akenhead) #4

Architecture. Continued Attribution.
Excellent man, that Tom Geudens! Here are some challenges.
(1) Need a sketch of required architecture and tools to take to Neo4j execs as the basis for creating a "sandbox" for ISDL. Out of my domain!
(2) Have datasets to share, but the biologists involved are keenly interested in "continued attribution" (needs a better term) so that products from their data maintain acknowledgement of who did what in the creation of this product. Can you (anyone?) suggest a model for this? Something about maintaining a link,
**(:Person{name:"Tom"}) **
**-[:Contributed{description:"Waist deep in freezing water, OK?"}]-> **
(:Dataset{label:"1993 Freezing Creek Coho Smolt Survey"})

If we make a plot of that data
(WTH? Now we have a workflow in Neo4j? with calls to R scripts? OMG.)
how does that link from Tom get duplicated and attached to a new node, something like
(:Media{label:"time series plot of Tom's data", URL="") ?

(Scott Akenhead) #5

My GraphConnect talk had a picture of a bear catching a salmon; people liked it. Encouraging me is a mistake, here's another. This is how nutrients from the ocean are distributed throughout a watershed.