Keywords

1 Introduction

We are on the way to an information technology society where there is a tendency to digitize information of all kinds in order to make it accessible to users. However, we are still a long way from the ideal of a uniformly digitized society in which everyone can use information without barriers. Access to information and communication is still closely linked to language skills  [2]. Many groups of people, e.g. people with learning difficulties, the ones affected by prelingual deafness, but also people with aphasia, are denied comprehensive participation in the socio-economic activities of society because they cannot read and understand standard language texts. In order to provide these people with unhindered access to information resources, barrier-free communication is required. The easy-to-read approach, which has experienced an upswing in recent years as a result of inclusion movements, is regarded as a central building block for creating this barrier-free environment  [10]. The easy-to-read approach aims at simplification at all levels of the source language with the aim of providing easily readable and comprehensible information. The transformation of standard language texts into easy-to-read texts can be regarded as an intralingual translation process, since no language boundary is crossed, but rather a variety boundary within a single language  [13]. While numerous computer-supported tools are available for interlingual translation processes, the translation of standard source texts into easy-to-read ones is primarily intellectual, which is why so far only a few high-quality easy-to-read texts are available. In order to be able to provide more information resources as easy-to-read with constant quality, the intralingual translation process needs to be optimized by means of adequate computer-supported tools.

Approaches that deal with the optimization of the intralingual translation process look in particular at the sub-process of text simplification. Initial approaches to text simplification, such as Chandrasekar et al.  [6], Caroll et al.  [5], Siddharthan  [18], investigate the grammatical complexity of a text with the aim of transforming it into a simpler structure and thus aim at syntactical simplification. Later work focuses on simplification at the lexical level by adding information from dictionaries or other sources  [8, 14] or by substituting difficult words  [4, 19]. Recent works, such as Van den Bercken et al.  [3], Nisioi et al.  [16], Paetzold and Specia  [17], focus on automated lexical text simplification using machine learning methods. To optimize the entire intralingual translation process, however, automation and digitization of further process steps are required. Nevertheless, approaches that deal with process automation and process digitization focus in particular on workflows in the industrial sector  [15, 20] and can only be used to a limited extent for the specific workflow of the intralingual translation process. The present study takes up this problem and examines which technical, qualitative and quantitative characteristics a software system must have in order to digitize the entire intralingual translation process according to demand and thus make it more effective, efficient and user-oriented.

2 Methodical Approach

Requirements form the basis of any software architecture. Systematic management and a structured approach to requirements analysis are fundamental to the development of efficient and error-free systems, as they ensure that the software system developed meets the needs of users  [1, 12]. For the determination, analysis, specification and validation of the properties and framework conditions of the required software system, we used the following systematic procedure, which is divided into three processes:

Process 1 – Define Scope of Requirements: The first process aimed to clearly define the scope of requirements. To this end, we analysed the intralingual translation process through extensive literature research in online databases (e.g. ACM, IEEE, Elsevier/Science Direct), conducted interviews with translators and practiced the intralingual translation process under the guidance of experts. In addition, we examined the process environment by looking at the intralingual translation process as a black box and identifying relevant areas of influence through systematic analysis.

Process 2 – Identification of Stakeholders and Primary Actors: Based on the analysis of the environment carried out in the first process, we then detected the stakeholders with the aim of identifying their requirements for the systems to be created. Since not all stakeholders were equally relevant, we prioritized the detected stakeholders according to their influence on and interaction potential with the system.

Process 3 – Requirements Assessment: This process focused on the assessment and specification of requirements. For systematic determination, a distinction was made between functional, non-functional, qualitative and technical requirements. The requirements were collected using two sub-processes (SP) described below.

SP 1 – Survey of Functional Requirements: To determine the functional requirements, we adapted the systematic approach of use-case analysis  [7, 11]. From the stakeholder survey, we generated use cases, which we converted into use-case diagrams.

SP 2 – Determination of Qualitative and Technical Requirements: In determining the qualitative requirements, we were guided by the international standard ISO/IEC 25000 “Software engineering - Software product Quality Requirements and Evaluation (SQuaRE)”  [9].

3 Results

The complex processes of the intellectual translation process from standard source texts into easy-to-read text were made transparent by a process and environment analysis and presented in a model-based activity diagram. Figure 1 shows the results of the process and environment analysis. The activity diagram consists of four sections, which are described as activities below: 1) Analysis of the standard source text, 2) Development of the translation strategy, 3) Preparation of the translation and 4) Verification of the translation. The individual activities are discussed below.

Fig. 1.
figure 1

Process activity diagram

Full size image

The first section is the analysis of the standard source text of the intralingual translation process. The analysis focuses, among other things, on the text length, the type of text as well as the function and information structure of the source text. If contradictions are found or information is missing, the translator establishes contact with the client. Based on the results of the first section and any further specifications, the translator develops a translation strategy for the source text in section 2. This strategy is tested for effectiveness by translating a small extract of the standard source text into an easy-to-read text. A direct feedback channel to the client can ensure that the intralingual translation is carried out in line with expectations. Section 3 then translates the entire standard source text into an easy-to-read text. The basis for this is the translation strategy, which determines how to proceed with the source text at word, sentence, text and character level. In Section 4, the easy-to-read text is then checked on the basis of a catalogue of criteria previously agreed with the client. If a target group correction is applied, the translation is forwarded to target group readers who check the easy-to-read text for comprehensibility and provide the translator with appropriate feedback. If the client wishes to have the text checked by a specialist, appropriate actions are taken before the text is forwarded to the client for final approval.

The results of the requirements analysis are described below based on the process shown in Fig. 2. In the first step, a system vision was developed based on the activity diagram and the essential system goals of the software to be developed were described. By means of the environment analysis we identified the primary, secondary and tertiary stakeholder groups. The primary stakeholder groups of the system to be created were translators, clients and easy-to-read readers. In addition, the group of translators was identified as the primary actor, as these are in many cases the triggers of a use case due to their direct interaction with the system in the application context and are therefore of particular importance for system creation. Developers of the software system could be identified as a secondary stakeholder group. We identified data center operators and system administrators as tertiary stakeholder groups. The third step involved modelling the system context. Based on the process environment analysis, the framework conditions of the intralingual translation process were first systematically analysed and then examined to what extent a software system could support this process and how it could be used in a targeted manner. It became clear that the support potential goes beyond the actual translation, i.e. the transformation of standard language texts into easy-to-read texts. For example, we identified significant potential in the analysis of the standard source text and at the interfaces between translator and client, which could make the entire process more effective and efficient in the form of direct feedback channels.

Fig. 2.
figure 2

Process of requirements analysis

Full size image

The implementation of steps one to three enabled the parallel execution of analysis steps A-E, the results of which are presented below: (A) From the activity diagram created and the survey of all stakeholders, in particular the primary actors, requirements for the software system were determined and specified by an iterative procedure. The functional requirements could first be formulated in natural language using a requirements template and then recorded in tabular form in requirement lists. In the interview procedure, goals and characteristics were worked out and precisely described (B). It became clear that all stakeholders were looking for a cross-platform solution, with a general increase in interest in digitization. This was mostly due to expected increased efficiency. The creation of use case diagrams (C) made it possible to visualize the findings recorded in the interview. These were used to check existing requirements for correctness iteratively with the stakeholders and to determine further requirements, in particular technical and qualitative ones. In this context, it became apparent that the stakeholders consider the scalability and maintainability of the software system to be particularly important. In a status quo analysis, we were able to determine the technical framework conditions of the stakeholders. It became clear that the majority of the stakeholders surveyed use desktop PCs with older Windows versions (Windows XP, Windows 7) and have little technical affinity. (C) Based on the findings, interface prototypes were developed using interactive mock-ups, tested together with stakeholders and the resulting feedback directly implemented. From this, it was also possible to identify missing requirements. It turned out that the users need a complex role concept and that the meta information for texts to be uploaded should also be adaptable and extendable for technically less experienced users. (D) A common glossary was created to ensure that all participants could use the same terminology. This was necessary because of the interdisciplinary nature of the work and the fact that certain terms may have different meanings in different fields. On the basis of all collected findings, it was possible to create a comprehensive final analysis document, which was optimized in several iteration phases together with the stakeholders and forms the basis for the development of a first prototype.

4 Conclusion

In our study, we conducted a systematic requirements analysis with the aim of identifying the qualitative and quantitative characteristics of a software system that are required to digitize the intralingual translation process according to the needs. To this end, we analysed the complex processes involved in the intellectual translation of standard source texts into easy-to-read texts and set out the process sequence in an activity diagram based on a model. By means of the activity diagram and the survey of stakeholders and primary actors, we conducted a systematic requirements analysis. We generated comprehensive requirement documents for the functional, non-functional, technical and qualitative requirements. In our future work, we will use the requirement documents created for the prototype development and implementation of a software system that can adequately support the previously intellectually executed translation process from standard source texts into easy-to-read texts. In this context, it is also necessary to check the identified requirements on the basis of test cases and, if necessary, to specify them further. This enables iterative optimization of the complex process.