Accommodation of Data Requirements Analysis to the Philological Context of the CIRCE Proje

Abstract

The problems derived from the collaboration between representatives of thematically diverse disciplines are a continuous focus of debate in the digital humanities field (Pitti, 2008: 485-486; Edmond, 2016: 54-56). Due to the need for designing, building, and administering a database containing philological information and a web application that gives access to said database, the philological research project “CIRCE: Early Modern Theatre on Screen” (https://circe.uv.es) has encountered the described issue since its inception. A fundamental objective when designing a software system with such characteristics consists of guaranteeing a requirements eliciting process whose results contemplate the team’s research needs, especially regarding the scope and structure of the data to store and manipulate. Said process must be managed including all stakeholders, in this case the project’s principal investigator, its members (researchers in the philology area who will also perform as the application’s users), and the information technology specialists in charge of its technical side. The success of such a task depends on adapting the linguistic register with which data needs are described to the abstraction level of the users (Sommerville, 2016: 103). Previous experiences in software development prove that the lack of collaboration between said actors aids the failure of such effort (Pressman and Maxim, 2020: 57-58; Harrington, 2016: 28).

To prevent misunderstandings in the identification of data requirements between the philological and technical members of CIRCE, a pre-analysis phase has been performed with the objective of determining which are the relevant data entities for the project’s purpose, which are the attributes of said entities to take into account, and which the accepted values for each attribute. Said activity has consisted in the elaboration of a series of data templates which have been distributed to the members of the project (future users of the web application), the inclusion by said members of information about audiovisual adaptations of European Renaissance theatre, and the subsequent analysis of the usage each team member has given to the templates. For the users of the application, the templates have played the part of use scenarios (Sommerville, 2016: 18); for the technical team, they have provided drafts of the data model to implement (Hills, 2016: n. pag.; Meier and Kaufmann, 2019: 25). The process has been refined iteratively through successive meetings and has allowed for an analysis of the project’s data needs whose level of detail and precision would have been hardly achieved had not all project members participated in it. This course of action has guaranteed the inclusion and implication of the whole team, mostly philological researchers, in a technical task generally associated with the software engineering discipline.

This article describes and explains said process, emphasising its interdisciplinary approach, its unifying attitude, and its interest in guaranteeing a high level of quality in the data requirements detection process. To achieve all of this, it references literature from different fields (digital humanities, software engineering, and database design), it remarks the need to apply software engineering concepts in similarly-scoped projects, it stresses the significance of attention to detail in data analysis (Date, 2019: 393), and it involves the user as an integral part of an information system’s data modelling process. The study of said process’s application to the context of the CIRCE project demonstrates the advantages of avoiding haste in the identification of data requirements thanks to assigning a generous amount of time to said task and enforcing the involvement of the users. The presence of mixed professional profiles that combine knowledge from such dissimilar spheres as philology and computing is assessed as advisable, as it reduces the communication problems derived from the absence of a common language between members of different knowledge domains.