Service-oriented Architectures (SOAs) for the Humanities: Solutions and Impacts

Large research infrastructure projects in the Humanities and Social Sciences such as Bamboo (http://www.projectbamboo.org/), CLARIN (www.clarin.eu), DARIAH (http://www.dariah.eu/), eAqua (http://www.eaqua.net/index.php), Metanet (http://www.meta‐net.eu/), and Panacea (http://www.panacea‐lr.eu/) increasingly offer their resources and tools as web applications or web services via the internet. Examples of this kind include:

• Bamboo Technology Project (http://www.projectbamboo.org/infrastructure/)
• eAqua Portal (http://www.eaqua.net/portal/)
• Language Technology World Portal of MetaNet (http://www.lt‐world.org/)
• PANACEA platform (http://www.panacea‐lr.eu/en/project/the‐platform/)
• TextGrid – eScience methods in Arts and Humanities (http://www.textgrid.de/en.html)
• VLO – Virtual Language Observatory (http://www.clarin.eu/vlw/observatory.php)
• WebLicht – Web Based Linguistic Chaining Tool (https://weblicht.sfs.uni‐tuebingen.de)

Such web-based access has a number of crucial advantages over traditional means of service provision via downloadable resources or desktop applications. Since web applications can be invoked from any browser, downloading, installation, and configuration of individual tools on the user’s local computer is avoided. Moreover, users of web applications will be ensured to always use the latest version of the software, since it will be updated on the host computer. It is exactly this ease of use that is of crucial advantage for eHumanities researchers, since configuration and updates of software often require computational skills that can ordinarily not be expected from humanities researchers.

The paradigm of service-oriented architectures (SOA) is often used as a possible architecture for bundling web applications and web services. While the use of web services and SOAs is quickly gaining in popularity, there are still a number of open technology and research questions which await more principal answers:

• Currently, web services and SOAs in the Digital Humanities often concentrate on written material. Will the current technology scale up to accommodate multimodal data like speech or video data as well?
• Currently, web services and SOAs typically process data in a synchronous fashion. How can very large data sets such as multimodal resources be processed in an asynchronous fashion?
• Currently, web services and SOAs tend to deliver analysis or search results in a non-interactive fashion, allowing user input only to initiate processing and to react to the processing result. How can the current applications be extended so as to allow dynamic user interaction during processing? Such considerations are of crucial importance for the eHumanities in order to support, inter alia, interactive annotation of text corpora, a desideratum for all text-oriented disciplines such as literary studies, history, and linguistics.
• Will web-based access over time completely replace stand-alone (downloadable) desktop or CLI applications, or will there always be a need for both: local and web-based applications?
• What is the impact of emerging technologies such as web sockets or cloud computing on existing web service environments?
• Currently, SOAs tend to be application or domain specific, catering to the data formats and services most relevant to particular user communities. What are the possibilities for generalizing such current practice and developing generic execution models and standards?
• How to generate knowledge from data, e.g. developing new digital methods and concepts such as new and adapted data structures, hierarchical data storage, data modeling, sorting and search algorithms, selection of data via metadata, and visualization tools?

You can download the proceedings of the workshop as PDF file.