Meeting minutes
Logistics
endo: Endo from NHK
… due to COVID-19 case, joining from Tokyo remotely
Endo: Media content include TV programmes and news articles
… The reach of content is limited due to fragmentation of deliver methods or services and platforms
… Use cases for content distribution across industries, and hints for cooperation across industry
<nigel> Nigel Megitt, BBC, co-chair TTWG and ADCG
Chris Needham, BBC. Media WG co-chair, and MEIG co-chair
daihei: Daihei Shiohama, co-chair of Publishing BG
<igarashi> Tatsuya Igarashi, Sony Group, co-chair of MEIG, and AB member.
<atsushi> Atsushi Shimono, W3C
<ktoumura> Kunihiko Toumura, Hitachi
<shiestyle> Shinya Takami: KADOKAWA (publisher in Japan), co-chair of Publishing Maintenance WG
<Daihei> Daihei SHIOHAMA, Media Do International (eBook distribution in the US) co-chair of Publishing Business Group
<nhk-ryo> Ryo Yasuoka, NHK
<ohmata> Hisayuki OHMATA NHK
Kaz Ashimura, W3C
<Bert> Self-intro: Bert Bos, W3C, on the I18N WG, Math WG and SDW WG, and W3C communications team.
<Ege_> Ege Korkan, Siemens AG
<Mizushima_> I am Tomoaki Mizushima from Internet Research Institute, Inc.
Timo Kunkel, Dolby Labs
<pat> I am Pat Griffis/Dolby Labs
<oyoshiba> Osamu Yoshiba from. KODANSHA, Publishing House in Japan
<tomikura> I am Yukio Tomikura, from Kodansha, Japanese publishing company. I'm in charge of selling e-books
Session Overview
Endo: In this session we distinguish between media and non-media insdustry
Endo: media industry inludes broadcasting and OTT services and publishers
Endo: NHK has proposed various use cases where media content can be used across industries
… In previous TPAC sessions, 2022 a WoT breakout session, the need for common rules for media content metadata was noted
… So far, there have been effots on inter-industry collaboration, on media player functions, rights management technologies, controlling IoT devices, metadata vocabulaties
… Best practices for media content metadata is not sufficiently shared
… We have two media industry presentations
Endo: The first is from broadcasting, the second from publishing
… We want to discuss four main topics: efforts for cooperation in the media industry, coorperation between media and other industries
… we welcome any Q&A
Use cases from the broadcast industry
Ohmata: Current use cases and issues from the media industry in Japan
… from NHK and metadata providers, IPG and M Data
… Starting with current use cases and issues. Broadcast service was launched in Japan in 1953, and digital since 2000
… Typical way to access TV programmes is from a programme guide. Thinking about the programme data, it's multiplexed into the brodcast side, then demuxed and presented to viewers as an EPG
… The format of the data is a broadcast standard. In Japan, there's no platform operators for terrestrial
… All broadcasters now provide internet services. We've provided web pages for programmes and news since 2000, and catchup and live of TV programme services
… NHK provides content metadata to platforms. Other broadcasters also provide catch up programmes
… Conventionally broadcasters only had to deal with standard formats for broadcast, but APIs vary by platform, so the operational cost is too much
… Data exchange tools, and checking of data
… NHK issues, we're now renewing web pages to use schema.org to make it easier to find and link pages and programmes
… But operators aren't motivated to input the metadata
… By describing the data items necessary for all platforms, the size of the metadata is increased
… Use cases from metadata providers: TV programme and retail business, the media and non-media industry use case
… IPG is a metadata company which provides data from broadcasters, and add information to the metadata and organise related content
… Their vision is opportunities to reach content by maintaining the metadata
… When the user selects a programme, the services that can provide the program are presented
… To provide the service there's a deep link to the catchup video on each service
… When we search by programme name or actor in Google search, the metadata can be presented
… We use Google's knowledge panel format, to give more accurate information to viewers
… Issues include renumbering of continuous broadcast of episodes, different rules for structurnig VoD content varies by platform
… Another metadata company is M Data. They gather text data from all TV programs broadcast in Japan
… Two types of metadata: scene data, which station and time, and topic presented, what items were presented. Commercial data includes company name, URL of e-commerce site, price
… Broadcast has influence on daily life, so TV metadata has the potential to connect broadcast to other services
… Supermarkets can use TV metadata to promote sales
… There are issues, not easy to create metadata when the product in the TV program cannot be uniquely identified
… Mapping TV metadata to the data used in the retail industry isn't easy
… Now, retail industry tries to make common rules on the data formats
… To summaries, TV metadata must be created for each OTT platform, so it's neessary to discuss the design of metadata to coorerate across industries
… What to do to solve these issues? Align to common web standards?
… Active research area in NHK
Yasuoka: Proposed system from NHK R&D
… The content discovery system is a web based system for delivering TV programmes
… We created a prototype
… [shows demo video]
… We're also presenting use cases that connect to services other than broadcasting and OTT
cpn: Is the situation that you do not have a common metadata standard across all broadcasters
… and service providers that works across all devices?
… Which standards do you use if they exist? Are there standards that are not well adopted
… or are you trying to move from multiple competing standards to a harmonized approach?
Ohmata: There are broadcast standards, and the OTT service has many formats, so a minimum requirement is common metadata to connect services between broadcast and other services
Kaz: There are two levels of question: common or integrated metadata for broadcast purposes, then the next level is thinking about further development and integration of metadata for all media distribution related industries, not just broadcasting
… So which direction to go? Create an integrated standardised metadata, or think about how to reuse metadata from different industries
Endo: Our prototype for future use cases is based on schema.org. Other standards, i think the media industries shares the metadata for media content, and cooperate with other industries
Igarashi: I'd like to clarify, you'd like to create a new metadate standard for OTT, because they don't accept broadcast standards?
Endo: Yes, I agree
<pat> pat griffis in the queue
Endo: Existing industries would like to use existing standards or specs where possible
… many standards or data catalogs, we'll contribute to those groups
Igarashi: Do you expect to have a new standard that's shared across the broadcasters?
Endo: Operators includes broadcasters and OTT providers, they have common issues, what industry collaboration does this need, we identified from interviews
<Zakim> nigel, you wanted to ask if you are proposing a new standard or promoting adoption of an existing standard? Is there a protocol for pointing to broadcast content?
Nigel: Are there already standards for the protocols for reaching the linked content. The video shows someone clicking on a programme, goes to a broadcast or OTT. Is this something that exists already as a standard?
<igarashi> NHK is promoting schema.org, is not it?
Endo: Existing standards are possible. We don't propose a new standard. We can share a minimal common vocabulary or best practices to describe the metadata
<ohmata> Yes, NHK think schema.org is a better solution.
Endo: One example is that TV episodes are usually described according to the TV series and the TV seasons in addition to the TV episodes themselves. However, each TV programme or an OTT programme doesn't have such a structure, like schema.org
… so which parameter is used by some operators in each company is decided by themselves
… So my idea is we can share best practices
Pat: With my SMPTE hat on, the issue of uniquely identifying content on a global scale is something SMPTE does already with studio content, and Netflix
… So I'd encourage you to bring it to SMPTE to get a solution not only for broadcast but also pre-production
… The metadata is complex. Hollywood movies have 300 versions, multiple languages, different cuts, etc
… So if we're having a system to identify content, it needs to allow for other forms of content
… Another use cases is user-generated content? How to tag that so that it can interoperate?
… The problem of identifying content could be a SMPTE work item, to get other broadcasters involved
Daihei: I represent the media side of a publishing business
… The content has its own rules, they have to merge, but at the same time if SMPTE works on it, but the content side doesn't know
… On the media side we have difficulty on how to make them available, so some interface is needed
Kaz: Endo provided ideas about collaboration across industries. For example, ARIB is also working on metadata
… Getting information from related SDOs about their work on metadata and collaboration with media-related companies including publishers in addition to broadcasters would be great. And now, we should ask Daihei for opinions from Publishers' viewpoints.
Ideas from Publishers' viewpoints
Daihei: Publishing movies, prorgammes, manga, etc. Everything should be consumed on a smartphone. So many challenges, I divide them into the creator/supply side or the demand/user side and the service side
… There's a gap. New things, WoT and digital twins. For the time being we may have to wait, we shoulnd't tolerate that
… Content are generated and produced for a single purposed, movies, books for print, music. All those media formats are made to be enjoyed in a certain designed format
… Read a book from beginning to end, watch a movie, but now we have opportunities where people may only use a short segment, based on searching for particular facts or knowledge
… This may invite some media content that enhances what they're looking for
… Also mash-up type uses, integrating parts of what they like
Daihei: [reads slide: Legacy linear organic growth vs Supply/Demand sides meeting disruptively]
Daihei: This is quite high level, idealistic. But in my experience from Japan, US, Europe across the media and publishing industry, when new media comes out you need to get the source rights owners view
… There are existing stakeholders on each side, knowing SMPTE is looking at it is useful, someone should have an overall view
Wrap-up
Endo: To conclude, we have broadcasting and publishing use cases. Any suggestions for continuing the discussion and future plan?
Kaz: Given the interest in the topic, could we bring this idea to MEIG and continue the discussion there, and survey SMPTE, ARIB, schema.org work, and consider the use cases more
Endo: I agree, we do need more time to discuss
… Also collaborate with other industries outside the media industry
Igarashi: I agree to continue in MEIG. But we need to continue investigation, e.g., if SMPTE works on this. I'm not sure if it's a global issue or specific to Japan
… I'd like to establish that before starting in MEIG
… OTT services aren't in MEIG, which is why I suggest surveying them first
Ohmata: We could make a CG to allow others who aren't members to participate
Kaz: yes, creating a new CG is another possible option :)
Pat: ARIB, ATSC and HbbTV, many are working on this challenge, seeing if there's a collective interest makes sense
… Not all may be appropriate for W3C. Rights issue, SMPTE has a registry where you can tag what rights are available
… It'll take many parties working together. I support bringing the broadcast community together
Endo: I'm sorry I couldn't be with you today. Please continue the discussion, and join our next meeting.
[adjourned]