W3C

– DRAFT –
Media Content Metadata: Practical Approaches for Cross-Industry Collaboration

10 November 2025

Attendees

Present
alwu, Andreas_Tai, Chris_Needham, Daihei, Endo, Haruki, hirata, Hiroki, Hisayuki_Ohmata, jay, Junko, Kamata, Kaz_Ashimura, Kunihiko_Toumura, lei_zhao, naomi, shiestyle, Yuta_Hagio
Regrets
-
Chair
Hiroki Endo
Scribe
kaz, cpn

Meeting minutes

hiroki: Welcome to the breakout. I'm from NHK, and chair of the Media Content Metadata Japanese CG
… We'll discuss practical approaches to improve metadata interoperability, focusing on practical use cases
… We'll have some live demonstrations
… We look forward to your feedback
… This group was created following a breakout at TPAC 2023
… Our goal today is to report on progress on interoperability, and get feedback for future collobration
… I encourage everyone to share their perspectives
… [Reviews the agenda]

Introduction and background

hiroki: We had a breakout at TPAC2023 on media metadata interoperability. This showed the need for ongoing discussion based on practical challenges
… An example challenge presented, operational costs are increasing due to adapting data to each OTT platform's API or data format
… and data must be verified separately for every platform
… In that session, broadcasters, metadata providers, and publishers shared their issues
… To address the challenges, MCM-JP CG was created
… This session will present the outcomes of 1 year of CG activity, and invite discussion on future direction

Mission and scope of MCM-JP CG

hiroki: The mission is to promote interoperability of media metadata across industries
… We shared practical case studies, best practices, building a foundation for mutual understanding
… The CG's main activities are collecting scenarios from each industry, and sharing case studies, surveying global standards and initiatives
… We report industry knowledge, such as issues, solutions, and metdata
… And combining industry knowledge on global standards, report on the feasibility
… We have a survey of specifications related to media content metadata. Each specification is well established, and useful in its respective industry
… In practice, existing workflows are deeply integrated, and there is little motivation to move to a different specification unless there is a specific need

Focus areas

hiroki: The first focus is industry knowledge sharing. We shared more than 10 case studies, these are in the draft report
… The second focus, we gathered 6 use cases that use existing industry metadata
… Some scenarios can be implemented using existing industry metadata. They don't create new metadata or rules for interoperability
… 6 proof of concept demos show that some use cases are already possible today
… Demo 1: event information linked with book and TV program information
… This application provides event information, such as festivals. In prototype testing, integrates publishing metadata and TV program metadata
… This shows the feasibility of combining the metadata was confirmed
… Demo 2 shows how metadata interoperability can be applied in an automotive environment
… In vehicle infotainment system, using the NetFront Browser for Automotive, from Access browser vendor
… Information is combined seamlessly, across devices and contexts, both personal and in-car devices
… Demo 3 focuses on integrating broadcasting and publishing metadata, integrated with an Input Method Editor, a tool for entering non-Latin characters by combining keystrokes into characters
… It provides predictive text and autocomplete features
… The text predictions are based on the content metadata
… In the next example, there is an intentional incorrect spelling, Pet Semetary. Without publishing metadata, the IME would recommend changing the word
… so we allow input of the correct, accurate title
… Another example, if the user enters "conan", the system offers the official English title, "Case Closed"
… Demo 4 highlights how the IME dictionary helps the user find the right content. The user often makes mistakes on input, so the system provides accurate auto-correct suggestions
… The approach benefits users searching for content, and business use cases where accurate input is essential
… In this scenario, contributed by a publishing member, two industry companies collaborated on the use case
… Demo 5 shows how publishing promotions from a TV program. By integrating broadcasting metadata, such as title, broadcast date and time, and ISBN, the publishers can highlight a book that has been introduced on TV, making their recommendations more relevant and timely
… This was contributed by M Data, a TV metadata provider
… Demo 6 enables new cross-media experiences, e.g., TV programmes recommended from e-books, related to the content

Conclusion

hiroki: MCM-JP CG gathered use cases. and outcomes so far include industry knowledge collected into the CG report. and 6 scenarios identified
… Some scenarios do not yet reach best practices, so we continue to collect industry knowledge

Discussion

hiroki: How can we build on these results?
… We have prepared three topics to guide the conversation
… Topic A is the scope of practical collaboration areas. What is feasible today? Where do we see opportunities for joint work?
… Topic B is requirements for promoting initiatives and building partnerships. What is needed to build a stronger community?
… Topic C is scope of MCM-JP CG and role sharing with other W3C groups and external standards organizations

sho: I'm from Shueisha publishers. Information like manga titles can already be shared. More complex are character names and storylines. We need to organise the data before we can share it
… This experience shows what we can do now and what we can prepare for future collaboration

hiroki: Thank you, it's good to know the status

Kaz: The work so far is great, it would be nice to think about how to describe the mechanism in a generic manner
… The auto-complete example is interesting, and we could describe the concepts.
… The scope of the group is Japanese, but describing the mechanism more would be helpful for Japanese and international industries

hiroki: I agree
… It's a Japanese CG because we start with industry case studies in Japan, because media content of often created to target a specific domain
… Japanese media content is discussed, and now we do interoperability. We start with industry case studies, so international interoperability is future motivation
… The findings we share in MEIG, which I'm participating in. We collaborate with Chris as chair of MEIG
… As MCM-JP CG chair, I'm collecting industry scenarios, and try to clarify common requirements for future internationalised recommendations

Jay: Collaboration isn't so easy. Metadata semantics may be the same, but the format might be different depending on the organization
… For example, the given name or family name. These formats mean the same person, so how to understand that one value could appear differently in different formats
… Maybe AI can play a role
… How to explain these different formats, but having the same semantics? That's a possible area of collaboration

hiroki: That's an important aspect

<jay> ack

hiroki: AI can analyse the data, but it learns from existing data. M-Data is creating metadata using AI, so for metadata creation AI is often used
… We can more easily connect other vocabularies using AI. Model Context Protocol is often introduced as a future AI technology
… Have a vocabulary and vocabulary description. The data has structure, with detailed metadata. MCM-JP CG industry knowledge is converted to this example or vocabulary description
… One example for collaboration

Yukio: The challenge we face is that publishing metadata is not designed for collaboration with other industries. This CG could transform how we approach publishing metadata

hiroki: Some publishers have both long and short titles, which the IME needs to allow for

Topic C: Let's clarify the scope of MCM-JP CG and discuss how to coordinate with other standards bodies

Andreas: It could be interesting to look at DVB-I, they are working on content guides, a more accessible way with TV Anytime for TV and media services
… There could be a link with what you're doing, with the content guide

Hiroki: Japanese broadcasters only have metadata for EPG officially. M Data provides detailed metadata. This, and DVB metadata or other international metadata should be connected in future
… An accessible approach is important, and encouraging business potential is important

Next steps

hiroki: We identified the need to continue the activities, with MEIG, and Publishing BG

Demo showcases

hiroki: Please try out the demos!

[adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 248 (Mon Oct 27 20:04:16 2025 UTC).

Diagnostics

Succeeded: s/cccccbcthhcirldjjvuivdgrelllchtrukjtgiiniucj//

Succeeded: s/Demo 3/... Demo 3/

Succeeded: s/Shueshan/Shueisha/

Maybe present: Andreas, Kaz, sho, Yukio

All speakers: Andreas, hiroki, Jay, Kaz, sho, Yukio

Active on IRC: alwu, breakout-bot, cpn, Daihei, hagio_nhk, Haruki, hirata, hiroki, jay, jkamata, kaz, ktoumura, lei_zhao, naomi, ohmata, shiestyle