Quantitative Drama Analytics' (QuaDramA) is a mixed-methods project that brings together researchers from modern German literature and computational linguistics, thus belonging to the field of computational literary studies.The goal of the project is to define, annotate, automatically detect, and quantitatively analyze different dramatic character types in German-language plays.To this end, we extract textual and structural properties from plays and investigate their distribution among literary characters, such as Romeo and Juliet.One of the key decisions in any mixed-methods project is to agree on a specific collaboration workflow between the different parties, yet this decision is rarely made deliberately and explicitly.In QuaDramA, the collaboration is based on three pillars.(i) Text annotation is used to both clarify concepts and enrich data required in machine learning.(ii) The different project parts perform close and frequent personal collaboration, particularly in the beginning of the project.In addition, quantitative analyses are made by researchers from both disciplines by using generic programming tools (thus avoiding the need for custom graphical user interfaces).(iii) Quantitative and qualitative research methods go hand in hand and are closely integrated.We discuss the reasoning behind these pillars in this article and provide some recommendations for projects with a similar setup. 2

The digital transformation is accompanied by two simultaneous processes: digital humanities challenging the humanities, their theories, methodologies and disciplinary identities, and pushing computer science to get involved in new fields. But how can qualitative and quantitative methods be usefully combined in one research project? What are the theoretical and methodological principles across all disciplinary digital approaches? This volume focusses on driving innovation and conceptualising the humanities in the 21st century. Building on the results of 10 research projects, it serves as a useful tool for designing cutting-edge research that goes beyond conventional strategies.

Zusammenfassung This article systematically explores different methods to describe and compare the thematic content of character speech in drama. The problems here are two-fold: The determination of the theme of a text segment is challenging, as it might be subtle or deliberately hidden. Secondly, the evaluation of this determination is difficult, because gold standards do not exist. We approach these problems by comparing the output of different systems to previously stated expectations (postulates) that are partially derived from drama theory and history. Results show that all systems currently have difficulties reaching high performance numbers. Topic modeling and systems based on a general-purpose dictionary ( GermaNet ) achieve comparable performance.

Sharedtasksareaworkformatprevalentinthenaturallanguageprocessingandmachinelearningcommunity. Thisspecial issue continues the reporting on the shared task SANTA (Systematic Analysis of Narrative levels Through Annotation), which has the development of annotation guidelines for narrative levels as its goal. Narrative levels, also known as embedded narrations, are omnipresent in many kinds of narrations, and one of the core concepts of narratology. In this introduction, we summarize the current state, report on the second annotation round in SANTA, draw some conclusions and, finally, derive some recommendations for future shared tasks in the digital humanities.

This article deals with the operationalization of character types in German-language drama. Based on literary and theatre history, we identify characters that belong to one of the three types: ›schemer‹, ›virtuous daughter‹ and ›tender father‹. For the characters in our corpus, property-based data sets were established, which are then used in turn for their automatic classification. In addition to the complexity of characters and the theoretical determination of character types, we discuss the methodological challenge of generalizing from a small set of annotations. Our experiments show that the selected types emerge as definable subsets within a population.

This chapter presents the basic considerations, the evaluation scheme and the results of the first shared task in and for digital humanities.The shared task aims to create annotation guidelines for narrative levels and has been able to attract eight participating teams in 2018.The evaluation scheme combines the dimensions 'conceptual coverage', 'applicability' and 'usefulness' with the measurement of inter-annotator agreement and was able to show that it is possible to make complex humanities concepts intersubjectively applicable through annotation guidelines.The winning guideline of the shared task is characterised by finding a good balance between the different goals and requirements.

El desarrollo de métodos computacionales cuantitativos para llevar a cabo un análisis estructural y formal de textos dramáticos es relativamente reciente en las Humanidades Digitales. En este artículo estudiaremos, a través de este enfoque cuantitativo, algunos aspectos de la obra de Calderón de la Barca. Para llevar a cabo el análisis se hará referencia a las poéticas dramáticas contem-poráneas de la Comedia nueva española, que se contraponen a las reglas establecidas en el teatro clásico francés, como las unidades de acción, lugar y tiempo, las reglas de distribución de los personajes, la convención sobre la separación de los estados sociales y sobre la composición en general. Las características cuantitativas se explorarán en los dos géneros dramáticos más empleados por Calderón, la comedia y el auto sacramental, detectando así los patrones y estructuras formales del autor.

The article describes the idea, operationalization decisions and results of the first shared task in the Digital Humanities.In the task, different participating teams developed annotation guidelines for narrative levels independently.Annotation guidelines are a prerequisite for the development of systems for the automatic detection of textual phenomena, and thus needed in Computational Literary Studies because they allow large-scale studies of narrative phenomena.The developed guidelines were compared using a newly developed evaluation scheme that brings together the three dimensions of conceptual coverage, applicability and usefulness.

Prison writing is a literary genre that is little noticed in the German-speaking world, yet it has been and still is very popular in a global and historical perspective. So it is intensively examined by international researchers. This essay is concerned with a subset of this genre, the prison letter. On the one hand, texts of this genre are by definition subject to very similar production conditions; on the other hand, their form and function strongly depend on the diversity and transformation of penal systems as a condition of communication. This relationship will be explained in a systematic and historical perspective with regard to the aesthetics of existential writing.

zur Konferenz Digital Humanities im deutschsprachigen Raum 2020 Passive Präsenz tragischer Hauptfiguren im Drama

In this section, we will discuss our idea of guideline evaluation and the underlying considerations. Evaluating annotation guidelines in this way is a fairly new endeavor, and we have developed the evaluation setup from the ground up. Although we do not claim our choices to be universally valid or applicable, we believe that this approach to guideline evaluation is relevant for similar settings and can be adapted to projects that might have other preferences and priorities.

Within literary studies, there is a coexistence of di erent perspectives on protagonists, heroes or main characters in dramatic texts, which provide di erent de nitions and strategies for the identi cation of those characters. Essentially, most of these de nitions can be translated into a set of machine-readable character traits. Characters that correspond to these traits may then be classi ed as protagonists of the drama in question, and be distinguished from other characters (e.g. minor, secondary, supporting characters). Designing an applicable classi cation is the central objective of this article. Part of the problem lies in identifying eponymous characters, which is related to classifying protagonists, but involves its own presuppositions. We start by approaching both tasks from a theoretical perspective and suggest our own de nition of a protagonist, which can be operationalized for the purpose of machinable classi cation but still draws on existing research in literary studies and follows its de nitions. An attempt at manual annotation shows however that this type of de nition possesses only a limited potential for intersubjectivity. Using a variety of features such as token count of characters, topic modeling and network sizes, we then train a random forest classi er that separates characters into protagonists and non-protagonists or eponymous heroes and non-eponymous heroes, respectively. The results show that protagonists and eponymous heroes are in fact reliably identi able using simple features because of their usually prominent position within the play. Following the literary studies perspective, a conclusive analysis of the classi cation of speci c characters using the examples of Die Verschwörung des Fiesko zu Genua, Maria Stuart, and Emilia Galotti makes clear that machine learning models o er interesting starting points for more in-depth re ections on protagonists and eponymous heroes. Bibliogra sche Information der Deutschen Nationalbibliothek: Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliogra e. Sie ist in der Zeitschriftendatenbank (ZDB) und im internationalen ISSN-Portal erfasst. Detaillierte bibliogra sche Daten sind im Internet über http://dnb.d-nb.de abrufbar. Alle Rechte, auch die des auszugsweisen Nachdrucks, der fotomechanischen Wiedergabe und der Übersetzung, vorbehalten. Dies betri t auch die Vervielfältigung und Übertragung einzelner Textabschnitte, Zeichnungen oder Bilder durch alle Verfahren wie Speicherung und Übertragung auf Papier, Transparente, Filme, Bänder, Platten und andere Medien, soweit es nicht ŸŸ 53 und 54 UrhG ausdrücklich gestatten. c ©2018 Benjamin Krautter, Janis Pagel, Nils Reiter, Marcus Willand ISSN: 2629-7027 Ed. by Thomas Weitin LitLab Pamphlet #7: Character Classi cation Benjamin Krautter, Janis Pagel, Nils Reiter, Marcus Willand Eponymous Heroes and Protagonists Character Classi cation in German-Language Dramas

Annotation guidelines for literary phenomena are a clear desideratum within the field of text-oriented digital humanities. Creating guidelines that are widely applicable, however, is almost only possible in large annotation projects, which are naturally expensive.

zur Konferenz Digital Humanities im deutschsprachigen Raum 2019 Ein neues Format für die Digital Humanities: Shared Tasks. Zur Annotation narrativer Ebenen

This volume is the first of two, and it documents activities that we have been conducting in the past years. They are best described as “organizing shared tasks with/in/for the digital humanities” and have evolved significantly since we started.

In this chapter, we give a descriptive overview of the annotation guidelines, their use of narrative level concepts, and the results of their quantitative evaluation. We will also connect some of the results to qualitative findings we uncovered during the workshop, although some references to the participants’ objectives are conjecture. Finally, the chapter contains a reflection on the annotation and evaluation procedure.