# openWEMI meeting May 17, 2023
###### tags: `openwemi`

Hackmd link: https://hackmd.io/QnDJSmDMR2Ooln3KysBe1g

Join Zoom Meeting:
https://us02web.zoom.us/j/89101555665?pwd=QW83cDFpQ1pETjB6YUNQOTFiQmZjUT09

## Times:
* San Francisco (USA – California) 9:00:00 PDT UTC-7 hours
* New York (USA – New York) 12:00:00 EDT UTC-4 hours
* London (United Kingdom – England) 17:00:00 GMT UTC
* Berlin (Germany – Berlin) 18:00:00 CET UTC+1 hour
* Corresponding UTC (GMT) 16:00:00

[Check time](https://www.timeanddate.com/worldclock/fixedtime.html?iso=20230517T1600) in your time zone

## Last week's notes
* https://hackmd.io/GoU1iKESTc-867dS2kcRKw?both

## Participants

* Graeme, Ross, Sean, Karen

## Agenda
* confirm decision to drop `RelatedEntity` as a class
* see also [issue 13](https://github.com/dcmi/openwemi/issues/13)
* move issues [#33](https://github.com/dcmi/openwemi/issues/33) and [#30](https://github.com/dcmi/openwemi/issues/30) (music and articles) to examples area?
* Examples: [Comics](https://github.com/dcmi/openwemi/tree/main/examples/comics) and [Bibo](https://github.com/dcmi/openwemi/tree/main/examples/bibo)
* transitivity [issue 24](https://github.com/dcmi/openwemi/issues/24)
* issues to close?
* [Is subclass between WEMI and Endeavor enough?](https://github.com/dcmi/openwemi/issues/14)
* [Related but not responsible](https://github.com/dcmi/openwemi/issues/17)

## Minutes
* Ross: did BIBO example using a book that becomes a BBC series, also using BIBFRAME. Added videos as related works. BIBO is flat, not divided like WEMI. Would like to also find as schema.org. Archive.org puts out metadata that is odd at times, but we should model it since it's real data
* Sean: comics example; shows how he is modeling the comics vocabulary properties to use openwemi. Lots of different aspects (cover, ink, coloring, story).
* Sean asks Ross: do people really use inferencing? Ross: we don't. It's very logical/resource intensive. Inferencing is about you transforming someone else's data so you can ingest it. Sean: it means adding many more triples. Ross: reason to include correct
* drop class RelatedEntity
* kc: we don't need to work on "other responsibility entities"
* move music and articles into examples - ACTION:kc
* bibframe example with hub is difficult - Graemi: odd that hub is subclass of work; otherwise makes sense. Sean: seem to be treating hub as superwork. Graeme: you can give attributes to the hub rather than the work.
* kc: don't have to do things just one way. Use eamples to show different interpretations
* close issues 14, 17, 24 (which was wrong); 14 endeavor is "just a bucket" (Graeme). Does related entity need a higher level? kc: schema.org has, for example, funder. Do we want a direct/indirect connect to WEMI? All of the entity relationships we can think of could be considered 'related'. Graeme - define related entity broadly enough that we can include all of these relationships. Related entity is too vague because ends up including subject relationships.
* "Transitive" is the wrong question - what we need is "inverse". Yes to adding inverse. Discussion ensues as to the merits (and demerits) of having both directions as properties. Are they needed for search? Are they too much for input? Is this solved with OWL:inverseProperty? Sean: many vocabularies already had both so having both in openWEMI aligns with that. Karen: this complicates openWEMI because it gives everyone two different properties for each relationship. Some inverses are in dublin core but best to think about this in terms of the flatness of DC. Karen will create new issue (Done: [#40](https://github.com/dcmi/openwemi/issues/40)) and create diagram.
* next week: this, and commonX