| News | Blog | Interdisciplinary challenges on the way to public value-oriented software engineering

Interdisciplinary challenges on the way to public value-oriented software engineering

Making sure that software-intensive systems are compliant with regulatory requirements, in domains such as public service media, say, may not be the most straightforward task. Indeed, we face special challenges when it comes to the engineering of such systems for modern media organisations. These challenges range from the implementation of diversity as a legal concept in recommender systems to the basic question of what forms of personalisation lie within the legal remit.


Nowadays, the implementation of values in different types of software-intensive systems – from social media to AI – continues to be topical in many countries around the world. More and more legal frameworks are starting to emerge that focus specifically on the regulation of software, its underlying logics and its output; the latest example at the time of writing this post is the EU proposal for an AI Regulation. Besides its main focus being concentrated on the types of values to be implemented and ever-new approaches to formulate respective requirements such as diversity or transparency, there is also another less prominent aim related to the implementation of already existing requirements in the cross-cutting concern of ‘public value’.

In Germany, legislators have tasked public service media (PSM) providers with a wide remit, containing duties, inter alia, to inform, educate, entertain and integrate society.

§26 State Media Treaty [Medienstaatsvertrag]

(1) The mission of the public service broadcasters is to act as a medium and factor in the process of free individual and public opinion formation by producing and disseminating their offers and thereby meeting the democratic, social and cultural needs of society. In their offers, the public service broadcasters have to provide a comprehensive overview of international, European, national and regional events in all major areas of life. In this way, they should promote international understanding, European integration and social cohesion in the federal and state governments. Their offers are used for education, information, advice and entertainment. They have to offer contributions in particular to culture. Entertainment should also correspond to a public service profile.

(2) When fulfilling their mandate, public service broadcasters must take into account: the principles of objectivity and impartiality of reporting, the diversity of opinions; and the balance of their offers.

Beyond this general legal remit, the State Media Treaty also contains specific requirements for online services.

§30 State Media Treaty [Medienstaatsvertrag]

(3) The contemporary design of the telemedia offers is intended to enable all population groups to participate in the information society, provide guidance, offer interactive communication options and promote the technical and content-related media literacy of all generations and minorities. This design of the telemedia offer should take special account of the needs of people with disabilities, in particular in the form of audio description, provision of manuscripts or telemedia in simple language.

This form of service provision is expected to produce added value for society (‘public value’). The responsibility for realising this remit as well as appropriate legal provisions lies with the public service broadcasters (their pluralistically composed internal supervisory bodies being the central player). As software systems are becoming an essential part of modern media organisations, it becomes increasingly expected that the legal framework is also about ensuring that such value-driven requirements are seamlessly reflected by public service media platforms.

Interdisciplinarity wanted

This situation poses several challenges. For instance, the engineering of public value-related requirements prescribed in media law is a complex and nontrivial endeavour: Such regulations are typically described in a normative, non-technical way, allowing for various technical interpretations (particularly when faced by engineers with limited legal backgrounds). Furthermore, the interpretation of what constitutes ‘value’ from a user perspective is open to legal interpretation, in other words diametrically different to common software engineering practices where the exploration and validation of user requirements (and what brings value to them) are inherently within the scope of engineering activities. Hence, such requirements can be adequately addressed only by interdisciplinary teams. When considering such interdisciplinary engineering team configurations from a superficial perspective, we would suggest a rather straightforward and linear interaction between disciplines, whereby inputs provided by legal experts would be processed by software requirements engineers, with such requirements then subsequently being implemented within media platforms. However, the real-world complexity renders this approach of little benefit and merely sustainable in practice. Such an instinctive, one-way, top-down approach is not helpful in the effective regulation of software systems.

The problem in a nutshell

The state of research on aspects related to regulatory requirements engineering for software systems reveals many ‘translation’-related problems. Requirements engineering methods are implicitly considered simple or even simplistic from a legal point of view or are not anticipated at all: Regulations that apply to software specifically usually use the same semantics and terms as provisions that apply to people. From a software engineering perspective, regulatory requirements are described vaguely and hardly implementable in an unambiguous manner. It is clear that the root of this problem lies mainly with different disciplinary perspectives being taken on the same issue of ensuring software systems’ compliance with any regulatory requirements, including those related to public value-orientation. A comparative analysis between legal and software engineering-specific approaches to public value-related regulatory requirements shows profound differences and suggests the reasons for challenges surrounding the effective implementation of regulatory requirements within software systems.

This, somewhat simplified comparison is supported by early results from our research project, where we study the state of practice versus the state of the art. There are no documented systematic approaches to implement regulatory requirements and ensure the compliance of software systems to regulatory norms in a seamless and reproducible manner. Instead, it seems that available requirements engineering approaches still suggest an improvised interaction between legal and domain experts on the one side with software and requirements engineers on the other side. Moreover, while some of the existing approaches to regulatory requirements engineering suggest ways to elicit the required legal and domain knowledge in a limited way, they do not allow for a sustainable and effective two-way interaction.

The approach taken in the ‘Coding Public Value’ project

Because of the contemporary challenges discussed above, in our project we are ambitiously seeking to tear down disciplinary boundaries related to developing engineering approaches to bring about more compliant requirements engineering of value-driven, public service media platforms. Here, we are aiming at effectively integrating: software engineering research; legal research; science and technology studies; together with media and communication research. These are all perspectives which are called for in the seamless implementation and assessment of existing public value requirements, contributing to actually coding public value using two key strategies:

Firstly, we are aiming to assist software engineers constructively through the seamless integration of regulatory requirements with software requirements engineering practices and integration/incorporation of the legal knowledge required in software engineering.

Secondly, we want to ensure that public value requirements are implemented in an auditable and verifiable manner for regulatory compliance purposes (either internally in PSM or externally, for instance when it comes to public accountability). At the level of our project, research results will update existing research on values and regulatory compliance in software engineering with a pragmatic requirement engineering methodology applicable for PSM-specific regulatory requirements of German media legislation. At the level of the interdisciplinary research and work, our results will create new opportunities for not only better regulatory compliance of software generally in areas of vague legal terms, but also simultaneously better consideration of software domain requirements: Let legal expertise and software requirements engineering ‘speak’ to each other.

The blogs published by the bidt represent the views of the authors; they do not reflect the position of the Institute as a whole.