This late breaking proposal is based on an ongoing effort started in the CLARIN project and which was presented briefly at the LREC 2012 Workshop on Collaborative Resource Development and Delivery. Initial point was the development of an easy to use information system for the description of standards developed in ISO/IEC TC37/SC4 Language Resources Management. Since these standards are heavily related to each other it is usually not feasible to adopt only a single standard for one's work but to dive into the standards jungle in full. Because of positive feedback after the presentation of the prototype at the Istanbul Workshop we decided to broaden the information accessible by the system to markup language related standards.
The problem with standards
Every year at Balisage's markup game or quiz difficult questions regarding markup have to be answered. Even the participants of this very special conference have sometimes problems to find the correct answer. An some of the younger generation have never heard terms like 'sosofo' or 'DSSSL' before. At current, an unmanageable number of standards is available for annotating data of various kinds. These standards can be divided into several groups, according to such different features such as standard body (W3C, ISO, OASIS, HL7, to name just a few), basic or derivative work (i. e., standards that are built upon basic specifications), the state (de jure vs. de facto standard), or the topic. Figure 1 gives a very rough overview about some of the named aspects. Other divisions would deal with a temporal aspect of standardization: specifications change over the time of development, some (almost) historical standards have been abandoned and have been replaced by other specifications.
Missing in the aforementioned list is another interesting aspect: the relationships between different standards. For example, meta languages such as SGML or XML are used to define (that is, syntax and possibly a schema formalism) markup languages. A given markup language is defined by a schema which in turn is defined by using a schema language (either a grammar-based or rule-based constraint language), and so on. Apart from these relations between basic standards and those that are built upon these, are other relations between members of the first group and of the second group. Some of the formal restrictions of XML instances and XML DTDs (and even XSD) are based on faits accomplis created during the development of SGML, DSSSL and HyTime – standards that are decades old. Features that were already present in DSSSL, have been improved and adapted for current W3C standards such as XPath, XSLT and XSL-FO. XPath is usually used via a host language such as XSLT. The Corpus Encoding Standard (CES) is an application of the SGML-based version of the TEI, P3. The are dozens of similar relations, some are only good for get bonus points at Balisage Bingo but others really help in understanding specific issues one may have when working with a given specification. At least the knowledge about these older standards and the decisions that were made during their design process and which still impact current technologies such as XML schema or XSLT runs the risk of being forgotten just because of the amount of time already passed.
This holds especially when dealing with non-basic standards, for example with markup languages for annotating linguistic data. During the last couple of years a large number of specifications have been developed in ISO/IEC TC 37/SC 4 Language Resources Management. These standards are often released to the public during their various stages of the standardization process (either as publicly available specialisations, such as Draft International Standard or as topic of a research paper). However, these version may substantially differ from the final version of the standard which is usually not available without charge. In addition, the relations between these specifications are numerous which complicates their correct use.
Another problem is the fact, that often scholars and researchers are not even aware that standardized formats and models exist.
We propose a community project to build up a platform providing guidance through the jungle that has been grown around the XML world. Starting with a very small set of standards and specifications and constructed as an XRX (XForms, Rest, XQuery) application we offer the starting point for a platform that allows Balisage's experts to share their knowledge with others. During the last months we have developed a prototypical web-based information system that serves as a starting point in providing guidance through the standards jungle as part of the CLARIN distributed project group. Up to now, it contains a collection of topics such as Meta Language, Metadata, Generic Corpus Annotation, or Constraint Language, amongst others, standard bodies such as ISO, W3C, OASIS, and HL7, and 25 specifications at the time of writing. Figure 2 shows a partial screenshot of the current state. Relations between specifications (in this case between TEI P3 and SGML and P3 and CES) are described both in a textual and graphical way.
The information system itself is based on standards as well. We have developed a lightweight format for structuring information about specifications, topics, and standard setting bodies defined by an XSD. Although it seems to be at least questionable to invent yet another annotation format especially for this kind of project we tried to stick as close to existing annotation formats such as the TEI as possible while streamlining the format and therefore keeping it small and simple. Figure 3 shows an excerpt of the storage format.
The information is stored in a native XML database system (we have chosen the Open Source eXist database as starting point but try to do not use any application-dependant features). Queries on the data are performed via XQuery scripts and forms will be implemented by XForms (which in turn will be processed by XSLTForms supported by eXist). The goal of the information system is not to replicate information that is already available (but may not be traceable anymore), but to connect pieces of information and enrich these pieces with small amounts of additional data. In the example above (Figure 3) we only refer to the information available at the W3C (and respective places). Information about standard bodies and topics is stored in a similar way.
Current state and future work
The current state of the information system is still quite rough. What we want to propose is allow Balisage's participants to get involved in the process of collecting and sharing information about the standards they work with and have knowledge of. In addition, Balisage is the place to find experts in SVG (which is used to display relations between two or more specialisations), XQuery and XForms. If we manage to bring these people together to establish a community that is willing to share its knowledge the final product could be of much use to scholars and researchers around the world. As a first starting point we will publish the XML schema defining the annotation format to receive comments and add further enhancements. After a stable format has been established, interesting parties could create specification sheets and upload them into the platform. In addition, we will open the platform for reading access for other people to give feedback on a less technical way. The current prototype is made available at http://clarin.ids-mannheim.de/standards/index.xq.
[Bostock et al., 2011] Michael Bostock, Vadim Ogievetsky, and Jeffrey Heer. D3: Data-driven documents. IEEE Trans. Visualization & Comp. Graphics (Proc. InfoVis), 2011.
[Jettka and Stührenberg, 2011] Daniel Jettka, and Maik Stührenberg. Visualization of concurrent markup: From trees to graphs, from 2d to 3d. In Proceedings of Balisage: The Markup Conference, volume 7 of Balisage Series on Markup Technologies, Montreal, 2011. doi:https://doi.org/10.4242/BalisageVol7.Jettka01.
[Stührenberg et al., 2012] Maik Stührenberg, Antonina Werthmann, and Andreas Witt. Guidance through the standards jungle for linguistic resources. In Nancy M. Ide, Collin Baker, Christiane Fellbaum, and Rebecca Passonneau, editors, Proceedings of the LREC 2012 Workshop on Collaborative Resource Development and Delivery, pages 9–13, 2012.