Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

MovieLabs publishes common ontology for production

ibc.org

Post reply on HN
Print view View original

21–30 of 33 posts

Re: MovieLabs publishes common ontology for production

#21

Post by jmole »

nerdponx wrote: post #17
danachow wrote: post #6

Earlier quoted context omitted.

Between ISO, UN CEFACT/OASIS, the now irrelevant/repivoted OMG, etc it isn’t like the idea is a new one or that anything isn’t written down - there are formal standards (insert xkcd here). The incentives for widespread adoption are just not there.

Maybe the problem is too much generality? An industry-specific ontology seems a lot more tractable than a general purpose ontology for all industries.

And cost - most standards organizations charge money for implementations and for training. I find it no coincidence that the most widely implemented standards are the ones that you can learn about and implement for free.

Re: MovieLabs publishes common ontology for production

#22

Post by alex_c »

ok123456 wrote: post #7

Are there any real success stories for "ontolgies" like this? It seems like a lot of them are made, but they're ultimately not that useful. I've been forced to use some, and it seems that someone got paid a lot of money to come up with something that's incomplete and poorly models the domain. So the net change is some XML doodads that no one really wanted or needed.

Music metadata (Musicbrainz, AllMusic, etc...) comes to mind. The data is usually more important than the ontology itself, and the ontology is not really consistent across the entire industry, but it's a somewhat relevant example.

Re: MovieLabs publishes common ontology for production

#23

Post by ok123456 »

alex_c wrote: post #22
ok123456 wrote: post #7

Are there any real success stories for "ontolgies" like this? It seems like a lot of them are made, but they're ultimately not that useful. I've been forced to use some, and it seems that someone got paid a lot of money to come up with something that's incomplete and poorly models the domain. So the net change is some XML doodads that no one really wanted or needed.

Music metadata (Musicbrainz, AllMusic, etc...) comes to mind. The data is usually more important than the ontology itself, and the ontology is not really consistent across the entire industry, but it's a somewhat relevant example.

The scope of those is pretty limited (i.e., helping people categorize their discogs with a dozen or so common fields) and isn't really a "musical ontology" where you're hooking it up to a first order logic solver and reasoning about music or whatever mythical use case drives most of that work.

Re: MovieLabs publishes common ontology for production

#25

Post by alex_c »

ok123456 wrote: post #23
alex_c wrote: post #22

Earlier quoted context omitted.

Music metadata (Musicbrainz, AllMusic, etc...) comes to mind. The data is usually more important than the ontology itself, and the ontology is not really consistent across the entire industry, but it's a somewhat relevant example.

The scope of those is pretty limited (i.e., helping people categorize their discogs with a dozen or so common fields) and isn't really a "musical ontology" where you're hooking it up to a first order logic solver and reasoning about music or whatever mythical use case drives most of that work.

Sort of... They are used at scale to share and ingest data between services, for things like recommendations, etc. But it is messy, a lot of the ontology is more or less the same but every company/database does things slightly differently.

Most of the difficulties are with matching artists and content across different databases, which is a separate issue. The ontology itself is tangential, but one example where it's important (even if it's not a fully / cleanly solved problem).

Re: MovieLabs publishes common ontology for production

#26

Post by SkeuomorphicBee »

alex_c wrote: post #22
ok123456 wrote: post #7

Are there any real success stories for "ontolgies" like this? It seems like a lot of them are made, but they're ultimately not that useful. I've been forced to use some, and it seems that someone got paid a lot of money to come up with something that's incomplete and poorly models the domain. So the net change is some XML doodads that no one really wanted or needed.

Music metadata (Musicbrainz, AllMusic, etc...) comes to mind. The data is usually more important than the ontology itself, and the ontology is not really consistent across the entire industry, but it's a somewhat relevant example.

Unfortunately the music field never had such collaborative effort to build a unified ontology, so each big player have its own ontology, and they are not very compatible. From personal experience, building software that consumes data from multiple sources in music is a nightmare, it is impossible to design a data model that fits all the incompatible ontologies.

Re: MovieLabs publishes common ontology for production

#27

Post by alex_c »

SkeuomorphicBee wrote: post #26
alex_c wrote: post #22

Earlier quoted context omitted.

Music metadata (Musicbrainz, AllMusic, etc...) comes to mind. The data is usually more important than the ontology itself, and the ontology is not really consistent across the entire industry, but it's a somewhat relevant example.

Unfortunately the music field never had such collaborative effort to build a unified ontology, so each big player have its own ontology, and they are not very compatible. From personal experience, building software that consumes data from multiple sources in music is a nightmare, it is impossible to design a data model that fits all the incompatible ontologies.

Agreed… and it’s too late to fix it now.

Relevant XKCD: https://xkcd.com/927/

Re: MovieLabs publishes common ontology for production

#28

Post by marcosmr »

ok123456 wrote: post #7

Are there any real success stories for "ontolgies" like this? It seems like a lot of them are made, but they're ultimately not that useful. I've been forced to use some, and it seems that someone got paid a lot of money to come up with something that's incomplete and poorly models the domain. So the net change is some XML doodads that no one really wanted or needed.

There are lots of examples of successful ontologies in the medical and biomedical domains https://bioportal.bioontology.org/

Re: MovieLabs publishes common ontology for production

#29

Post by beaconstudios »

satyanash wrote: post #2

This looks like a large scale domain modelling problem for the movie production process. This is the last place I was expecting an overlap between computer engineering and film-making. But it is not surprising given that large parts of film-making are now supported by software, which has a tendency to direct the evolution of terminology and concepts. After a cursory glance at the model, it starts with a fairly generi…

Ontology is part of philosophy, and concerns itself with what exists; domain modelling is ontology applied to systems (typically software and/or bureaucratic systems).

Re: MovieLabs publishes common ontology for production

#30

Post by kbenson »

andyjohnson0 wrote: post #12
ok123456 wrote: post #7

Are there any real success stories for "ontolgies" like this? It seems like a lot of them are made, but they're ultimately not that useful. I've been forced to use some, and it seems that someone got paid a lot of money to come up with something that's incomplete and poorly models the domain. So the net change is some XML doodads that no one really wanted or needed.

> Are there any real success stories for "ontolgies" like this? It seems like a lot of them are made, but they're ultimately not that useful. Back in the dot-com era I was marginally involved in a project proposal for a business spun-out of a logistics company. They had developed an ontology that supposedly captured their parent business's (admittedly extensive) knowledge of global logistics, and had then built a mul…

Sometimes it's hard to know if a business failed because the idea was bad, the market wasn't ready, they didn't execute well, they didn't advertise themselves effectively to let clients that could benefit know what they could offer, or some combination of multiple of those.

What you described sounds really interesting and useful if I understand it correctly, and I can think of multiple times over the past decade plus that it would have been useful to use as a consultant for companies I've worked for, if it wasn't too costly.

Post reply on HN