Keywords

1 Introduction and Theoretical Background

Grounded Theory (GT) has proven to be an extremely useful research approach in several fields, including computer science; however, it is a complex method based on an inductive paradigm that is fundamentally different from the traditional hypothetico-deductive research model [1]. Constructivist grounded theory, a qualitative research methodology primarily developed by Kathy Charmaz [2, 3], emphasises the co-construction of knowledge between the researcher and participants. Unlike traditional grounded theory, which seeks to uncover an objective ‘truth’ or reality, the constructivist approach acknowledges that both parties bring their own experiences, backgrounds, and perspectives to the research process. This reciprocal interaction results in a richer and more nuanced understanding of the phenomenon being studied. One of the key benefits of constructivist grounded theory is its flexibility and adaptability, which allows researchers to delve deeply into complex social processes and understand them from multiple viewpoints. Additionally, it promotes reflexivity, encouraging researchers to critically examine biases and preconceptions throughout the research process. Charmaz's work highlights the importance of viewing research participants as active agents in the construction of knowledge rather than passive subjects, leading to more authentic and comprehensive insights. The principal distinctions between grounded theorists and other qualitative researchers have historically centered on the matter of preconceptions. Preconceptions hold great importance and influence researchers’ analysis. Despite remaining inconspicuous on the surface, a profound bond with research participants can ignite the researcher's passion and drive [4].

Grounded theory is ideal for examining sociocultural and human dimensions in software engineering, offering fresh perspectives. Although software engineering presents certain complexities for grounded theory studies, it continues to be a robust approach for theory construction and software requirement analysis. Developing a solid theoretical foundation is a significant hurdle in software engineering [5]. Grounded theory studies can contribute to the development of comprehensive theories and guide future empirical research in software engineering [6].

This paper proposes a new adaptive constructivist grounded-theory approach for Extended Reality Applications requirements elicitation (ACGTRE). As a test case, “Digicult_Aitoloakarnania. Augmenting Actual Visits and Cultural Tourism of the Fortification Architecture of Aitoloakarnania with the holistic use of Interactive Virtual and Augmented Reality Applications.”—project code: Τ6ΥΒΠ-00476 which co-financed by the European Regional Development Fund of the European Union and Greek national funds through the Operational Program Competitiveness, Entrepreneurship and Innovation, under the call Special Actions in aquaticfarming- industrial materials- open innovation in culture, was utilized.

Despite promotion, protection, and restoration, fortifications are not linked. To enhance digital visualization, three apps were developed: a virtual tour, virtual reality app, and augmented reality app. An adapted Grounded Theory was used to identify the apps’ requirements. Constructivist Grounded Theory was used to co-create knowledge and conclusions. This approach provides a novel way to understand Aitoloakarnania's fortifications and can be adapted to extended reality. It yielded “grounded” requirements and user stories. User stories from grounded requirements gave a clear understanding of user needs and expectations. This enabled the creation of applications that were both technologically advanced and user-centric, ensuring the final product was both useful and enjoyable. User stories connected the technical development team and end-users, ensuring the applications met user needs and expectations, improving the user experience. The team also ran “Virtual Street Museum—StreetMuseum” without Grounded Theory, creating an app with personalized cultural tourist routes. The app uses AR to enhance the user's experience, enabling them to save personal city data. It plans routes based on weather and user preferences. The project involves creating an algorithm for optimal routes, big data analytics, and AR mobile apps.

2 Methodology

Requirement analysis is a demanding process that involves many challenges. For this reason, a methodology which included many meetings between project team members for brainstorming, focus groups, and recording discussions for additional analysis was followed. The focus was then on the visitor's experience in fortifications, the description of his experiences, and what interests him in such a visit. Methods which use visitor-generated visual content (VGVC), like travel photos uploaded to platforms such as Flickr, offers insights into tourists’ behavior beyond traditional, costlier methods like surveys [7]. In addition, an important factor is the study and research in literature as well as the utilization of the experience of the project team. All collected data were analyzed using the inductive qualitative research method of Grounded Theory. The constructivist grounded theory method was used as the most appropriate method for which researchers and participants co-constructed the final knowledge and conclusions during analysis and data collection. The key questions of this study were: (a) Is there a need to select content categories specific to fortifications in terms of application development?, (b) How will applications meet the needs of end users beyond the basic level of technologies initially described from the perspective of other stakeholders (Ephorate of Antiquities of Aitoloakarnania and Lefkada)? For the modelling and visualisation of the requirements extraction process, the Business Process Modeling Notation. A diagram in the BPMN divides the process of the proposed adaptive constructivist grounded theory (ACGT) into four main stages:

  • Stage 1—Gathering Information.

  • Stage 2—Implementation of Methodological Approaches

  • Stage 3—Coding

  • Stage 4—Create user stories—Record requirements

2.1 Stage 1—Gathering Information

Two groups were identified for the project: internal (project team members) and external (app users and fortification visitors). Sources of information were identified for focus groups (mainly project team members) and external sources (mainly visitors). Original proposal, archaeological material, Google Maps and TripAdvisor reviews of Aitoloakarnania fortifications were selected for analysis.

2.2 Stage 2—Implementation of Methodological Approaches

At this stage, various methodological approaches were used to gather the material to be analyzed and codified in the next stage of the project. More specifically:

2.2.1 Interviews—Meetings

The meetings that took place were mainly online, and those related to wider meetings were recorded and transcribed in an automated way. The meetings can be divided into three categories: technical team meetings with 5–7 attendees, content creation team meetings with 4–6 attendees, and cooperation meetings with 8–10 attendees. All attendees are members of the project team. Automatic transcription, although not entirely accurate, combined with the video in the form of subtitles provides a thorough analysis of the meetings. In the context of this research, two meetings attended by the project team’s main members were used.

2.2.2 Focus Groups—Writing Texts

The focus groups worked on and/or created a series of documents following meetings between the software development team and the archaeology team. The research focused on creating texts concerning fortifications in various municipalities of Aitoloakarnania.

2.2.3 Comments of Visitors to Fortifications

As part of the search for information from visitors to the fortifications of the region of Aitoloakarnania, comments were collected from the two most popular platforms, TripAdvisor and Google Maps. More specifically, comments were collected about the fortification monuments for which there were apparent comments registered on the platforms. For the above sites, 90 pages of user comments were collected.

2.3 Stage 3—Coding

The coding stage involves a series of sequential actions of initial coding, focused coding, and theoretical coding. At the same time, throughout the successive actions, the actions of continuous comparison of codes, creation of notes in the form of memos and creation of code categories run—constant comparative analysis [8].

2.3.1 Initial Coding

In the initial coding phase, we explore any theoretical possibilities from the data. This helps us decide on basic conceptual categories. Comparing data reveals participants’ issues, which we can then address in detail. The coding follows the data closely. We observe actions on each data point instead of using pre-existing categories. The aim is to code with words that show action. This coding method restricts conceptual leaps and premature application of theories. Wider meetings were transcribed, imported into ATLAS.ti [9], and coded accordingly. The project began with coding the proposal, then the focus groups’ texts. This was to study Aitoloakarnania's fortifications. Visitors’ comments from Google Maps and TripAdvisor were also coded. Before coding, word frequency was analysed with a word cloud and a table of words frequency.

2.3.2 Focused Coding

Focused coding is the second step in coding. It is more precise, selective, and conceptual than word-by-word, line-by-line, and incident coding. After initial coding, focused coding synthesizes and interprets larger data sets. Focused coding uses essential and/or frequent codes to analyze large data sets. It assesses their adequacy and requires decisions on which source code best categorizes data thoroughly. Focused coding is not a linear process. It involves selecting the most frequent or important codes to categorize data. The following categories were created as part of focused coding: documentation digitization tool, project applications, VR Virtual Tour App: PC Virtual Tour Application, Augmented reality app, monument entities, Spotlight approach, what concerns the visitor?

2.3.3 Theoretical Coding

Theoretical coding is an advanced form of coding that builds on focused coding. Glaser proposed theoretical codes to connect noun codes and form assumptions into a theory. In essence, theoretical codes identify the connections between categories from focused coding. The theoretical codes unify and shape the collected focused codes, creating analytical, coherent narratives and advancing analytical history. Correlating the basic categories (technical, visitor needs, and monument entities) generated grounded conclusions in the form of narratives structured as basic user stories, outlining high-level requirements. Key user stories are used to create more detailed user stories. These narratives will follow the format “As a… I want to… So That…I will achieve this through…” and will be based on data from stakeholder meetings, archaeological/historical and technical approaches, and visitor feedback. The study thoroughly analyzed the app’s features, visitor needs, and requested information, reaching a satisfactory level of saturation.

2.4 Stage 4—Create User Stories—Record Requirements

At this stage, having developed the established conclusions in the form of narratives, utilizing the experience of the development team in combination with data visualization (Sankey diagrams, network diagrams, etc.) extracted from ATLAS.ti [9], user stories were created based on the subsystems that emerged from the initial design and during the study. They were captured in tables as shown in Table 1.

Table 1 Template of user stories—requirements
Full size table

The objective was to develop narratives that would interpret central stories based on and supported by the analyzed data. These central stories could be characterized as “Initiatives” and then analyzed into “Epics” and “User stories”. The terms “initiatives”, “epics”, and “user stories” refer to elements of the agile methodology, which is very popular in software development. It is important to note that these specifications will be improved, and, in some cases, new specifications will be added as part of the project development cycles.

3 Results and Discussion

Both projects, Virtual Street Museum and Digicult_Aitoloakarnania, used a hybrid Agile/Stage-Gate methodology. This methodology supports iterative cycles and external collaboration in product development. This study focuses on the initial phases, particularly the requirements analysis. Constructivist grounded theory can improve project reliability and quality compared with traditional requirements elicitation methods. It also supports the visualization of the results (diagrams) and is compatible with the Agile Manifesto principles [10], as described in Table 2.

Table 2 Comparing Agile Manifesto with Constructivist grounded theory with corresponding visualization
Full size table

Agile initiatives are a modern approach to software development that focus on continuous improvement and adaptation to change. An important element of Agile is the use of “epic stories” and “user stories”. Epic stories are big tasks that require a lot of time and effort to complete, whereas user stories are smaller functions that the user wants to achieve with the software.

4 Conclusion

The requirements for the extended reality applications of Digicult_Aitoloakarnania were gathered using an adapted version of Grounded Theory, a qualitative research method that is inductive in nature and visualised using Business Process Management Notations. The Constructivist Grounded Theory method was employed to facilitate the co-construction of final knowledge and conclusions by researchers and participants during the analysis and data collection stages. This innovative approach offers a novel way of engaging with and understanding Aitoloakarnania's unique fortifications, and can be tailored to the design needs of extended reality applications. The outcomes of this approach resulted in well-defined and ‘grounded’ requirements, which led to the creation of clear and comprehensive user stories. These user stories, derived from grounded requirements, provide a thorough understanding of users’ needs and expectations. This clarity in user stories facilitated the development of technologically advanced and user-centric applications, ensuring that the final product was both functional and engaging. The user stories served as a crucial connection between the technical development team and end-users, ensuring that the developed applications were aligned with the users’ needs and expectations, thereby enhancing the overall user experience. During the same chronological period, a second research program with the name “Virtual Street museum—StreetMuseum” having the same main core of the development team, was running without using the adapted version of Grounded Theory. The methodology described above supports the Digicult_Aitoakarnania development team so that it has a quicker and more focused description of the requirements than the description from the Virtual Street Museum. Both approaches resulted in successfully developed applications, but in Digicult_Aitoakarnania, fewer agile iterations were required, resulting in a shorter development time. Our results are promising and should be validated in more projects. Automation of the methodology using artificial intelligence tools should also be the focus of future studies to improve the process and offer new insights of the data.