• DARIAH EU
• Open Access close
• Publications close
• Research data close
• Other research products close
• European Commission close
• EC|H2020 close
• EU close

This deliverable builds on and further extends the findings of D6.1 "Inventory of existing data sources and formats" surveying the landscape of literary corpora, as well as D8.1 "Tools for NLP" cataloguing the set of tools in the context of CLS. Focusing on the wealth of formats used when encoding and processing text, it offers a comprehensive overview of common formats for encoding textual data, beyond the "lingua franca", TEI, both in the domain of computational literary studies and computational linguistics, highlighting potential discrepancies in the approach between these two areas of research. The overview reveals a very heterogeneous landscape with a plethora of formats, devised for differing tasks, from philological encoding of historical text material, to computational annotation and processing of text. Considering interoperability an indispensable key to reusability, the deliverable explores the challenges and approaches converting between formats. This information compilation is considered input for further developing the Transformation Matrix, introduced in D6.1, which shall serve as a conceptual framework to consolidate existing solutions for format conversion in the Transformation Toolbox to be delivered by the end of the project (D6.2). The Transformation Matrix shall allow to capture information about specific data structures (features) present in datasets as well as data structures required or produced by tools. This requires a sufficiently expressive formalised description, which is proposed in the CLSCor data model.

While the calls for reproduction studies in Computational Literary Studies (CLS) have become louder, practical aspects, especially the interplay of the components involved in the research process (code, data, environments, infrastructures, etc.) proves to be a hurdle for reproducing research. We present a way to fully reproducible research using a Docker-based approach: We exemplarily implemented it for a network-analytic study on a corpus of about 3,000 theater plays derived from the DraCor project. We demonstrate that the use of highly portable, self-contained digital artifacts (Docker images) containing runnable research environments not only allow for a full reproduction of the study, but also offer ways to implement different scenarios of repeating research (e.g. same code, different data).

This deliverable presents three case studies involving digitisation and transformation processes; the studies are presented in order of the complexity of the research question, which is reflected in the difficulty of the corpus compilation task. Transformation processes seem to be inevitable in each case, but paradoxically the urgency of digitisation diminishes as the complexity of a task increases, The case studies described in this deliverable are: 1. Creation of an ELTeC affine corpus of the Slovak novel (chapter 2) 2. Finding the haiku across multilingual corpora (chapter 3) 3. Measuring entropy and surprisal in the prose of the Tsarist Empire Devoted to Terrorism (Russian and Polish Texts) (chapter 4) The first two case studies have already served as reference cases for the data landscape review (CLS INFRA Deliverable 5.1). This extended version, which conveys the experience of six months of research and is enriched by the third case study, highlights specific aspects of the multidimensional landscape of literary text collections. In Deliverable 5.1, they were merely illustrations and concretisations of general points; now they are the focus of attention. The third case has been designed with the most complex research questions in mind, to go even further in exploring what is available and what is possible in the digital humanities today.

The Skills Gap Analysis, a recent survey among scholars describes the current distribution of (self-attested) practical text-processing skills and scholars' despair from the necessary skillset, as well as a perceived absence of "an entry point in(to) CLS education". We argue that corpus analysis is the entry point and why.

While the calls for reproduction studies in Computational Literary Studies (CLS) have become louder, practical aspects, especially the interplay of the components involved in the research process (code, data, environments, infrastructures, etc.) proves to be a hurdle for reproducing research. We present a way to fully reproducible research using a Docker-based approach: We exemplarily implemented it for a network-analytic study on a corpus of about 3,000 theater plays derived from the DraCor project. We demonstrate that the use of highly portable, self-contained digital artifacts (Docker images) containing runnable research environments not only allow for a full reproduction of the study, but also offer ways to implement different scenarios of repeating research (e.g. same code, different data).

The TEITOK REST API lets you interact with your corpus remotely - to upload documents in various formats - automatically converted to TEI/XML; to render search results; to run an NLP pipeline on the server; or download the content of corpus document, treat that content locally with local NLP tools or manual annotation tools, and then upload the results back to the server, where the new or corrected annotations will be incorporated into the original TEI/XML document without destroying any of the potentially complex annotations already in the document.

The multilingual DraCor platform (https://www.dracor.org) represents a valuable resource for literature and theatre scholars, allowing them to host, access and analyse thousands of plays from Antiquity to the XX century. After briefly presenting the workflow for the ingestion of new plays into our ecosystem, we focus on the collaborative side of our endeavours, demonstrating how external scholars can benefit from a range of tools and guides to easily prepare and submit their own collections. As a showcase of the process, we present three corpora currently in production, focusing respectively on Ukrainian, Czech, and Early Modern English literature.