This was actually an issue raised by Mo McRoberts [1] in the context of his Digital Public Space work [2]. But I seem to have become his secretary :-/
In the programmes ontology [3], we have the concept of an episode (radio or TV). An episode can have one or many versions (longer, shorter, audio described, british sign language etc). These roughly correspond to the media assets that are broadcast / made available ondemand. Each version can have 0, 1 or many broadcasts and 0, 1 or many ondemands (representing vod / aod availability). Broadcasts are a type of event
Mo is has got some data about programmes that have been recorded from broadcast. And the mo:Recording event felt like a good fit with that
Which means we're missing a predicate between po:Broadcast and mo:Recording
There's also the question of whether a recording of a broadcast represents a new version of the episode (because it might have top and tail information (trails, squeezed credits etc). So are we missing another predicate between po:Version and mo:Recording
Or this version is a result of that recording of that broadcast of that other version of the same episode....
Does this work? Do we need new predicates? If so where should they live? Mo or po?
http://www.bbc.co.uk/ This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. If you have received it in error, please delete it from your system. Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. Please note that the BBC monitors e-mails sent or received. Further communication will signify your consent to this.
> This was actually an issue raised by Mo McRoberts [1] in the context of his > Digital Public Space work [2]. But I seem to have become his secretary :-/
> In the programmes ontology [3], we have the concept of an episode (radio or > TV). An episode can have one or many versions (longer, shorter, audio > described, british sign language etc). These roughly correspond to the media > assets that are broadcast / made available ondemand. Each version can have > 0, 1 or many broadcasts and 0, 1 or many ondemands (representing vod / aod > availability). Broadcasts are a type of event
> Mo is has got some data about programmes that have been recorded from > broadcast. And the mo:Recording event felt like a good fit with that
> Which means we're missing a predicate between po:Broadcast and mo:Recording
Perhaps you can use the generic event ontology for that, which would mean using event:factor to link a mo:Recording and a po:Broadcast - would that make sense? Anything else should probably be a subproperty of that.
> There's also the question of whether a recording of a broadcast represents a > new version of the episode (because it might have top and tail information > (trails, squeezed credits etc). So are we missing another predicate between > po:Version and mo:Recording
That's much trickier. In the programmes ontology, a lengthened/shortened programme would correspond to a different version - I guess it is probably the same case here?
On a side-note, one thing that is missing from the Programmes Ontology (but available in some form in the /programmes RDF) is the concept of a media asset - a version can correspond to many media assets (we don't recreate a version each time a media is reencoded for a particular device). If the format in which the encoding is being done is relevant, then it would be useful to have this concept in as well.
Also, would the event:product of the Recording be the version, or the media asset? (I'd go for the second one, I think).
> Or this version is a result of that recording of that broadcast of that > other version of the same episode....
> Does this work? Do we need new predicates? If so where should they live? Mo
I think the only thing that really needs adding is the media asset concept, which should probably live in PO. Although I know CME has been dealing with similar issues - so probably worth a look as well.
> http://www.bbc.co.uk/ > This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. > If you have received it in error, please delete it from your system. > Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. > Please note that the BBC monitors e-mails sent or received. > Further communication will signify your consent to this.
> -- > You received this message because you are subscribed to the Google Groups "Music Ontology Specification Group" group. > To post to this group, send email to music-ontology-specification-group@googlegroups.com. > To unsubscribe from this group, send email to music-ontology-specification-group+unsubscribe@googlegroups.com. > For more options, visit this group at http://groups.google.com/group/music-ontology-specification-group?hl=en.
(With thanks to Michael for writing up the guts of this while it languished on my to-do list!)
On 18 Aug 2011, at 15:05, Yves Raimond wrote:
> Perhaps you can use the generic event ontology for that, which would > mean using event:factor to link a mo:Recording and a po:Broadcast - > would that make sense? Anything else should probably be a subproperty > of that.
That could work… as far as that's concerned, it's two correlated events. I won't have too much data about the recording event itself, but there's a certain amount of implication between a mo:Recording and a po:Broadcast which is a factor in that event?
>> There's also the question of whether a recording of a broadcast represents a >> new version of the episode (because it might have top and tail information >> (trails, squeezed credits etc). So are we missing another predicate between >> po:Version and mo:Recording
> That's much trickier. In the programmes ontology, a > lengthened/shortened programme would correspond to a different version > - I guess it is probably the same case here?
That's my feeling.
Essentially I want to correlate the derivative version which represents the recording, with the broadcast (possibly via a mo:Recording --> po:Broadcast) and then a consumer can follow its nose up to the version which was played out.
> On a side-note, one thing that is missing from the Programmes Ontology > (but available in some form in the /programmes RDF) is the concept of > a media asset - a version can correspond to many media assets (we > don't recreate a version each time a media is reencoded for a > particular device). If the format in which the encoding is being done > is relevant, then it would be useful to have this concept in as well.
A predicate in the PO namespace for relating a po:Version to one or more media items would be lovely — do you happen to have a copy of some RDF which employs that form?
> Also, would the event:product of the Recording be the version, or the > media asset? (I'd go for the second one, I think).
Tricky. I'd have said the opposite… the version is created as a result of the recording, and is made available as one or more media assets. That also allows for descriptions of recordings which have since been lost.
Hmm, that solves the recording/version representation problem, then.
Something like this (once whatever “ex:hasMediaAsset” is settled upon)?
</programme/123456#version> a po:Version ; ex:hasMediaAsset </programme/123456/download/video-300k.mp4> .
(You could then describe </programme/123456/download/video-300k.mp4> as a concrete asset, including a po:OnDemand availability event?)
> I think the only thing that really needs adding is the media asset > concept, which should probably live in PO. Although I know CME has > been dealing with similar issues - so probably worth a look as well.
CME?
M.
-- Mo McRoberts - Data Analyst - Digital Public Space, Zone 1.08, BBC Scotland, 40 Pacific Quay, Glasgow G51 1DA, Room 7066, BBC Television Centre, London W12 7RJ, 0141 422 6036 (Internal: 01-26036) - PGP key 0x663E2B4A
http://www.bbc.co.uk/ This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. If you have received it in error, please delete it from your system. Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. Please note that the BBC monitors e-mails sent or received. Further communication will signify your consent to this.
Thinking about the concept of 'digital media' would there be a necessity of a digital asset ontology which related any digital representation of something (i.e. Actor). Not only could this relate to the recording and digital media from po & mo but could this relate to something like a digital version of sheet music thinking about the symbolic notation ontology?? This would also translate we to a museum perspective and their digital assets of objects???
On 18 Aug 2011, at 15:05, Yves Raimond <yves.raim...@gmail.com> wrote:
>> This was actually an issue raised by Mo McRoberts [1] in the context of his >> Digital Public Space work [2]. But I seem to have become his secretary :-/
>> In the programmes ontology [3], we have the concept of an episode (radio or >> TV). An episode can have one or many versions (longer, shorter, audio >> described, british sign language etc). These roughly correspond to the media >> assets that are broadcast / made available ondemand. Each version can have >> 0, 1 or many broadcasts and 0, 1 or many ondemands (representing vod / aod >> availability). Broadcasts are a type of event
>> Mo is has got some data about programmes that have been recorded from >> broadcast. And the mo:Recording event felt like a good fit with that
>> Which means we're missing a predicate between po:Broadcast and mo:Recording
> Perhaps you can use the generic event ontology for that, which would > mean using event:factor to link a mo:Recording and a po:Broadcast - > would that make sense? Anything else should probably be a subproperty > of that.
>> There's also the question of whether a recording of a broadcast represents a >> new version of the episode (because it might have top and tail information >> (trails, squeezed credits etc). So are we missing another predicate between >> po:Version and mo:Recording
> That's much trickier. In the programmes ontology, a > lengthened/shortened programme would correspond to a different version > - I guess it is probably the same case here?
> On a side-note, one thing that is missing from the Programmes Ontology > (but available in some form in the /programmes RDF) is the concept of > a media asset - a version can correspond to many media assets (we > don't recreate a version each time a media is reencoded for a > particular device). If the format in which the encoding is being done > is relevant, then it would be useful to have this concept in as well.
> Also, would the event:product of the Recording be the version, or the > media asset? (I'd go for the second one, I think).
>> Or this version is a result of that recording of that broadcast of that >> other version of the same episode....
>> Does this work? Do we need new predicates? If so where should they live? Mo
> I think the only thing that really needs adding is the media asset > concept, which should probably live in PO. Although I know CME has > been dealing with similar issues - so probably worth a look as well.
>> http://www.bbc.co.uk/ >> This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. >> If you have received it in error, please delete it from your system. >> Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. >> Please note that the BBC monitors e-mails sent or received. >> Further communication will signify your consent to this.
>> -- >> You received this message because you are subscribed to the Google Groups "Music Ontology Specification Group" group. >> To post to this group, send email to music-ontology-specification-group@googlegroups.com. >> To unsubscribe from this group, send email to music-ontology-specification-group+unsubscribe@googlegroups.com. >> For more options, visit this group at http://groups.google.com/group/music-ontology-specification-group?hl=en.
> -- > You received this message because you are subscribed to the Google Groups "Music Ontology Specification Group" group. > To post to this group, send email to music-ontology-specification-group@googlegroups.com. > To unsubscribe from this group, send email to music-ontology-specification-group+unsubscribe@googlegroups.com. > For more options, visit this group at http://groups.google.com/group/music-ontology-specification-group?hl=en.
foreword: albeit the Programmes Ontology doesn't follow the four levels of abstraction of FRBR, you'll derive them rather quick from the existing concepts, i.e. - po:Episode -> expression level - po:Version -> manifestation level - po:BroadCast, po:MediaAsset -> item level , or? ;)
On 8/18/2011 4:22 PM, Mo McRoberts wrote:
> Hi all,
> (With thanks to Michael for writing up the guts of this while it languished on my to-do list!)
> On 18 Aug 2011, at 15:05, Yves Raimond wrote:
>> Perhaps you can use the generic event ontology for that, which would >> mean using event:factor to link a mo:Recording and a po:Broadcast - >> would that make sense? Anything else should probably be a subproperty >> of that.
> That could work as far as that's concerned, it's two correlated events. I won't have too much data about the recording event itself, but there's a certain amount of implication between a mo:Recording and a po:Broadcast which is a factor in that event?
Please remember the Music Ontology deals specifically with _audio_ signals, i.e., the mo:Recording concept [1] cannot be used to represent TV show recordings. Furthermore, I would place such a recording more on the item (or manifestation) level than on the expression level, since I guess you would like to represent personal recordings of broadcasts, or? Besides, from my point of view the Programmes Ontology currently concentrates on knowledge representations of the broadcast domain - so one should keep the concepts and relations on this side of the information publishing cycle, i.e., the recording/consuming part is/can be represented with other, more appropriated ontologies, e.g., (if you are not constrained of modelling recordings) the Counter Ontology [2] (with its general ScrobbleEvent concept) or the Play Back Ontology [3].
>>> There's also the question of whether a recording of a broadcast represents a >>> new version of the episode (because it might have top and tail information >>> (trails, squeezed credits etc).
No, I think that such a recording doesn't represent a new po:Version of an episode, since we are leaving (at least for the moment) the broadcast part of the information publishing cycle (however, such a recording can be broadcasted again).
>>> So are we missing another predicate between >>> po:Version and mo:Recording
>> That's much trickier. In the programmes ontology, a >> lengthened/shortened programme would correspond to a different version >> - I guess it is probably the same case here?
> That's my feeling.
> Essentially I want to correlate the derivative version which represents the recording, with the broadcast (possibly via a mo:Recording --> po:Broadcast) and then a consumer can follow its nose up to the version which was played out.
So, you'll need a general, media-type-independent 'recording' concept. Furthermore, the products of mo:Recording instances are mo:Signal instances, i.e., I guess, you want to relate them (or their concrete encodings) to the broadcast or the broadcasted version of an episode.
>> On a side-note, one thing that is missing from the Programmes Ontology >> (but available in some form in the /programmes RDF) is the concept of >> a media asset - a version can correspond to many media assets (we >> don't recreate a version each time a media is reencoded for a >> particular device).
So, from my understanding 'media assets' represent po:Versions in different encodings, i.e., we are now on the item level ;)
>> If the format in which the encoding is being done >> is relevant, then it would be useful to have this concept in as well.
> A predicate in the PO namespace for relating a po:Version to one or more media items would be lovely do you happen to have a copy of some RDF which employs that form?
>> Also, would the event:product of the Recording be the version, or the >> media asset? (I'd go for the second one, I think).
> Tricky. I'd have said the opposite the version is created as a result of the recording, and is made available as one or more media assets.
I guess, we are leaving here the domain of PO for the moment (see above ;) ).
> (You could then describe</programme/123456/download/video-300k.mp4> as a concrete asset, including a po:OnDemand availability event?)
>> I think the only thing that really needs adding is the media asset >> concept, which should probably live in PO.
At the moment, I would also vote to include this 'media asset' concept into PO. However, then you are quite close (again) to the four levels of abstraction of FRBR (it's often so unavoidable, see, e.g., [4] - they are omnipresent ;) ).
>> Although I know CME has >> been dealing with similar issues - so probably worth a look as well.
> CME?
See [5,6].
I hope I could clarify some things for moment. Furthermore, I hope that my thoughts make more or less a bit sense ;)
> foreword: albeit the Programmes Ontology doesn't follow the four levels of > abstraction of FRBR, you'll derive them rather quick from the existing > concepts, i.e. > - po:Episode -> expression level > - po:Version -> manifestation level > - po:BroadCast, po:MediaAsset -> item level > , or? ;)
> On 8/18/2011 4:22 PM, Mo McRoberts wrote:
>> Hi all,
>> (With thanks to Michael for writing up the guts of this while it >> languished on my to-do list!)
>> On 18 Aug 2011, at 15:05, Yves Raimond wrote:
>>> Perhaps you can use the generic event ontology for that, which would >>> mean using event:factor to link a mo:Recording and a po:Broadcast - >>> would that make sense? Anything else should probably be a subproperty >>> of that.
>> That could work… as far as that's concerned, it's two correlated events. I >> won't have too much data about the recording event itself, but there's a >> certain amount of implication between a mo:Recording and a po:Broadcast >> which is a factor in that event?
> Please remember the Music Ontology deals specifically with _audio_ signals, > i.e., the mo:Recording concept [1] cannot be used to represent TV show
I disagree quite strongly on that :) There's nothing audio-specific to the mo:Recording event - the event could perfectly produce an audiovisual signal.
> recordings. Furthermore, I would place such a recording more on the item (or > manifestation) level than on the expression level, since I guess you would > like to represent personal recordings of broadcasts, or? > Besides, from my point of view the Programmes Ontology currently > concentrates on knowledge representations of the broadcast domain - so one > should keep the concepts and relations on this side of the information > publishing cycle, i.e., the recording/consuming part is/can be represented > with other, more appropriated ontologies, e.g., (if you are not constrained > of modelling recordings) the Counter Ontology [2] (with its general > ScrobbleEvent concept) or the Play Back Ontology [3].
>>>> There's also the question of whether a recording of a broadcast >>>> represents a >>>> new version of the episode (because it might have top and tail >>>> information >>>> (trails, squeezed credits etc).
> No, I think that such a recording doesn't represent a new po:Version of an > episode, since we are leaving (at least for the moment) the broadcast part > of the information publishing cycle (however, such a recording can be > broadcasted again).
>>>> So are we missing another predicate between >>>> po:Version and mo:Recording
>>> That's much trickier. In the programmes ontology, a >>> lengthened/shortened programme would correspond to a different version >>> - I guess it is probably the same case here?
>> That's my feeling.
>> Essentially I want to correlate the derivative version which represents >> the recording, with the broadcast (possibly via a mo:Recording --> >> po:Broadcast) and then a consumer can follow its nose up to the version >> which was played out.
> So, you'll need a general, media-type-independent 'recording' concept. > Furthermore, the products of mo:Recording instances are mo:Signal instances, > i.e., I guess, you want to relate them (or their concrete encodings) to the > broadcast or the broadcasted version of an episode.
>>> On a side-note, one thing that is missing from the Programmes Ontology >>> (but available in some form in the /programmes RDF) is the concept of >>> a media asset - a version can correspond to many media assets (we >>> don't recreate a version each time a media is reencoded for a >>> particular device).
> So, from my understanding 'media assets' represent po:Versions in different > encodings, i.e., we are now on the item level ;)
>>> If the format in which the encoding is being done >>> is relevant, then it would be useful to have this concept in as well.
>> A predicate in the PO namespace for relating a po:Version to one or more >> media items would be lovely — do you happen to have a copy of some RDF which >> employs that form?
>>> Also, would the event:product of the Recording be the version, or the >>> media asset? (I'd go for the second one, I think).
>> Tricky. I'd have said the opposite… the version is created as a result of >> the recording, and is made available as one or more media assets.
> I guess, we are leaving here the domain of PO for the moment (see above ;) > ).
>> That also allows for descriptions of recordings which have since been >> lost.
>> Hmm, that solves the recording/version representation problem, then.
>> Something like this (once whatever “ex:hasMediaAsset” is settled upon)?
>> </programme/123456#version> a po:Version ; >> ex:hasMediaAsset</programme/123456/download/video-300k.mp4> >> .
>> (You could then describe</programme/123456/download/video-300k.mp4> as a >> concrete asset, including a po:OnDemand availability event?)
>>> I think the only thing that really needs adding is the media asset >>> concept, which should probably live in PO.
> At the moment, I would also vote to include this 'media asset' concept into > PO. However, then you are quite close (again) to the four levels of > abstraction of FRBR (it's often so unavoidable, see, e.g., [4] - they are > omnipresent ;) ).
>>> Although I know CME has >>> been dealing with similar issues - so probably worth a look as well.
>> CME?
> See [5,6].
> I hope I could clarify some things for moment. Furthermore, I hope that my > thoughts make more or less a bit sense ;)
> -- > You received this message because you are subscribed to the Google Groups > "Music Ontology Specification Group" group. > To post to this group, send email to > music-ontology-specification-group@googlegroups.com. > To unsubscribe from this group, send email to > music-ontology-specification-group+unsubscribe@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/music-ontology-specification-group?hl=en.
>> Please remember the Music Ontology deals specifically with _audio_ signals, >> i.e., the mo:Recording concept [1] cannot be used to represent TV show
> I disagree quite strongly on that :) There's nothing audio-specific to > the mo:Recording event - the event could perfectly produce an > audiovisual signal.
Yes, reading the current definition of mo:Recording, there is no concrete association to audio-based recordings. However, since it is part of the Music Ontology, one usually associates audio-based recordings with this concept. Furthermore, mo:produced_signal [1] has the range of mo:Signal [2], which is a sub class of mo:MusicalExpression [3]. So again, a general, media-independent 'recording' concept is out of scope of the Music Ontology. It could be a part of, e.g., a media ontology.
>>> Please remember the Music Ontology deals specifically with _audio_ signals, >>> i.e., the mo:Recording concept [1] cannot be used to represent TV show
>> I disagree quite strongly on that :) There's nothing audio-specific to >> the mo:Recording event - the event could perfectly produce an >> audiovisual signal.
> Yes, reading the current definition of mo:Recording, there is no concrete association to audio-based recordings. However, since it is part of the Music Ontology, one usually associates audio-based recordings with this concept. Furthermore, mo:produced_signal [1] has the range of mo:Signal [2], which is a sub class of mo:MusicalExpression [3]. > So again, a general, media-independent 'recording' concept is out of scope of the Music Ontology. It could be a part of, e.g., a media ontology.
I'm not sure MTV would necessarily agree with that classification?
But in any case, one would use event:product (or some particular derivative) rather than mo:produced_signal in this case.
I'm not entirely convinced that reinventing an ontology with a set of mildly tweaked definitions to permit broader use would be the best solution to this class of problem…
M.
-- Mo McRoberts - Data Analyst - Digital Public Space, Zone 1.08, BBC Scotland, 40 Pacific Quay, Glasgow G51 1DA, Room 7066, BBC Television Centre, London W12 7RJ, 0141 422 6036 (Internal: 01-26036) - PGP key 0x663E2B4A
> I'm not entirely convinced that reinventing an ontology with a set of mildly tweaked definitions to permit broader use would be the best solution to this class of problem
Well, this unfortunately daily bread of knowledge representation and creating a 'common understanding' based on the definitions of the concepts. So it's not "reinventing an ontology", it's more "fitting the purpose" in a more appropriated (less inconsistent) definition/context. And it is (of course) representing/describing the reality (which is always complex somehow) ;) For example, the concepts of the Music Ontology can be derived from a more general media ontology, where concepts for a Video Ontology (oh, btw, there is now already one quite appropriated one there, see [1]) can be derived from this more general ontology as well.
Cheers,
Bo
PS: it's usually a matter of time to go not the whole way down and include as many abstractions as possible (which fit different contexts); however, at the end one probably satisfies a broader audience with different levels of abstraction/specialisation
a quick search on schemapedia.org helped me to rediscover known and existing things. In the 2008 Manu Sporny et al. implemented the bespoken modularization in a rather simple Media Vocabulary [1], Audio Vocabulary [2], and Video Vocabulary [3]. The Media Vocabulary has a concept called media:Recording. Maybe this one is appropriated to represent your TV and radio show recordings. You can utilized their specialisations as well - audio:Recording, video:Recording. What do you think about this solution?
>> I'm not entirely convinced that reinventing an ontology with a set of >> mildly tweaked definitions to permit broader use would be the best >> solution to this class of problem
> Well, this unfortunately daily bread of knowledge representation and > creating a 'common understanding' based on the definitions of the > concepts. So it's not "reinventing an ontology", it's more "fitting the > purpose" in a more appropriated (less inconsistent) definition/context. > And it is (of course) representing/describing the reality (which is > always complex somehow) ;) > For example, the concepts of the Music Ontology can be derived from a > more general media ontology, where concepts for a Video Ontology (oh, > btw, there is now already one quite appropriated one there, see [1]) can > be derived from this more general ontology as well.
> Cheers,
> Bo
> PS: it's usually a matter of time to go not the whole way down and > include as many abstractions as possible (which fit different contexts); > however, at the end one probably satisfies a broader audience with > different levels of abstraction/specialisation
