The objective of the community group is to propose a technology-neutral data model for electronic content labels, i.e. age labels or content descriptors. The data model will include agreed categories and fields that may contain content-specific information. The proposal is planned to include a documentation, code snippet examples and probable queries to support implementing the data model in existing age classification contexts.
The data model proposal and the documentation are planned to serve as guidelines for either existing players to implement the data model in their existing schemes (and thus providing users additional information in an interoperable way) or for new players that plan to label online content and thus reduce the risk of sunk costs.
Note: Community Groups are proposed and run by the community. Although W3C hosts these conversations, the groups do not necessarily represent the views of the W3C Membership or staff.
To consider views and comments from data and metadata experts as well as industry players in the field of age classification, the MIRACLE project currently conducts a public consultation on version 0.93 of its specification. The consultation has started on 25 June 2014 and runs for 4 weeks. You will find all relevant documents both on the MIRACLE website. All interested stakeholders are invited to comment on the proposed specification.
Right after the end of the consultation phase, the project members will discuss all comments received. If comments result in changes of the data model, the project will publish the respective version 1.0 shortly after. Feel free to dive into the proposal and don’t hesitate to comment on the specification. In case of any questions you might, of course, get in touch with the MIRACLE people.
Once the MIRACLE specification reaches v1.0, this community group will propose the specification in W3C format.
During spring 2014 a lot has happened: After reporting the outcome of the Task Force’s talks and discussions to the CEO Coalition to make the Internet a better place for kids, the participants of the Task Force finally agreed on a first proposal for a data model concerning interoperable, machine-readable age classification data. Please find attached the final proposal document.
However, for this W3C community group this is not the place to stop. If anything, this is the starting signal to work on a common technical specification for age classification information based on this proposal. How? Well, in the meantime the MIRACLE project has received funding by the European Union – MIRACLE stands for “Machine-readable and interoperable age classification labels in Europe”. The first thing of this pilot project will be to agree on a data model on the grounds of the attached Task Force proposal. The project consortium (with me being the project co-ordinator) will post the first draft of that technical specification here on the W3C platform in mid-June for public comments and consultation. In the meantime, have a look at the MIRACLE website for more information on the project and its activities:
Since the last entry, two meetings took place to discuss the data model proposal in detail:
30 Sep 2013: Recap of last meeting and status quo, discussion of specific data model issues, thinking ahead: implementation context and information exchange issues
9 Dec 2013: Agreement on final data set structure, proposed solutions for foreseeable implementation issues, recap of additional field proposals, dissemination and implementation opportunities in 2014
For the plenary meeting of the “CEO Coalition to make the Internet a better place for kids” on 24 January 2014, the Task Force on Interoperability and Machine-Readability prepared a report for the CEO Coalition members that sums up the starting point as well as the ongoing debates to a great extent (TaskForce 2013 Report).
After a year of discussions, we are currently in the final phase of agreeing to a first draft data model for interoperable age labels. Once we are ready the proposal will be uploaded right here.
Since a significant amount of discussions regarding an age label data model has taken place before the creation of this group, I’d like to give you an overview on the current state of play.
The Technical Task Force I mentioned here has met two times now, discussing a very rough working paper with options for a data model proposal that shows some of the lines of though behind the debates. The first on-site meeting took place in December 2012, the second one in April 2013.
Find attached the current version of a (yet to be discussed) proposal for an outline of how a common data model could look like. Since this work in progress you will see that there are still points that need to be addressed until a real proposal will be ready for distribution. Feel free to ask if anything’s unclear, and don’t hesitate to comment, of course.
Starting from the insight that the harmonisation of classification schemes is neither desirable nor feasible due to different socio-cultural contexts of classification, technical interoperability between the existing schemes and their electronic labels is seen as a possible way ahead to optimise the efficiency of machine-readable age labels and their comprised rating knowledge beyond national borders, to extend availability of online classification labels as well as to wider and more innovative ways of use of user-side information tools in general.
Main objective: A common data model for online labels
Before this background, this Community Group aims at proposing a technology-neutral data model for electronic content labels including agreed categories and fields that may contain content-specific information. The proposal is planned to serve as a guideline for either existing players to implement the data model in their existing schemes or for new players that plan to label online content and thus reduce the risk of sunk costs. The proposal is planned to include documentation, code snippet examples and probable queries to support implementing the data model in existing classification contexts. By doing so, the Community Group hopes to achieve better interoperability of classification data and electronic age labels in practice.
Basic principles of a data model
The data model has to build on currently existing practices, as it otherwise would undermine the efforts already taken by both companies and rating bodies as well as the classification knowledge that goes with such labels. For companies and bodies that already label online content electronically, no disadvantages should result from a proposal made.
The three basic requirements the data model will therefore take into account are:
The data model has to be technology-neutral to reach maximum openness and compatibility between different systems and languages.
It has to consider existing electronic labeling systems to ensure that these are not undermined by the interoperable data model.
It has to thoroughly take into account existing national and supranational classification schemes. By doing so, existing visual labels can easily be extended by respective electronic labels while at the same time ensuring compatibility with the data model.
One fundamental principle of the data model is that neither existing approaches and schemes nor future ones have to provide information in all categories – as long as the data that is provided by the label does fit into any of the categories, the system is technically interoperable. However, the more information a system or label provides, the better other systems will be able to use and process the data.
The relevance of interoperable labels for companies and users differ, depending on both the context of system environments (closed gardens: low relevance, open internet: high relevance) and the business case including the respective data transfers (B2B: probably lower relevance, B2C: high relevance). This results in a legitimate variance in company interest, while other relevant stakeholders (especially rating bodies, filter software providers, additional content providers, consumer associations, family associations) haven’t had the chance to participate in the discussions yet. Hence, this Community Group also aims at incorporating those bodies and/or their opinions in the discussion.
There are a lot of supranational, national, regional, media-, OS- and platform-specific classification systems and age labels out there, all aiming at informing children, their parents and other consumers about potentially harmful or unwanted content. The more these labels become digitized, the more potentials grow to extend the basis of consumer information by making existing labels (technically) interoperable.
Starting point for this community group have been considerations during discussion rounds within the European Commission-initiated “CEO Coalition to make the Internet a better place for kids“, namely within the Working Group 3 (“Wider use of content classification”). It has been agreed that to make classification data interoperable on a very basic level, we need to have a common reference model to exchange data and make it technically processable. A Technical Task Force on Interoperability and Machine-Readability has been created in autumn 2012 that convenes throughout 2013 to develop a proposal of such a data model.
The Task Force members agree to share the draft proposals here to extend the group of participating parties and to open the discussion to stakeholders and persons interested in interoperable age labels. Please join the group and briefly introduce yourself on the group’s mailing list: firstname.lastname@example.org (here’s the archive)