Der Band Fabrikation von Erkenntnis. Experimente in den Digital Humanities ist eine Gemeinschaftspublikation der Zeitschrift für digitale Geisteswissenschaften und Melusina Press. Die folgenden Beiträge werden im Teilband 1 durch Melusina Press publiziert und stehen Ihnen hier zur Verfügung.
Teilband 1
Using Deep Learning for Emotion Analysis of 18th and 19th Century German Plays
Long Paper
von Thomas Schmidt, Katrin Dennerlein, Christian Wolff
Abstract:
We present first results of the project »Emotions in Drama« in which we explore the annotation of emotions and the application of computational emotion analysis, predominantly deep learning-based methods, in the context of historical German plays of the time around 1800. We performed a pilot annotation study with five plays generating over 6,500 annotations for up to 13 sub-emotions structured in a hierarchical scheme.
The Sensitivity Topic Coherence Measurement to Out-of-Vocabulary Terms
Long Paper
von Keli Du und Steffen Pielström
Abstract:
Topic modeling is a popular method in computational literary studies (CLS). For its evaluation, topic coherence measures have been proposed that quantify the semantic coherence of topic words by analyzing their co-occurrences in a larger reference corpus. As a consequence, they can not produce meaningful results for words that do not appear in the reference corpus.
Disruptionen der Literaturwissenschaft am Beispiel der DVjs. Methodische Validierung durch Simulation und Anwendung
Long Paper
von Leonard Konle, Fotis Jannidis, Steffen Martus
Abstract:
This article deals with the use of distance measures to detect disruptive phases in literary studies using the Journal DVjs. In this context, the impossibility of methodological validation by annotated data is problematized and an alternative way of evaluation by simulation is proposed. Further, we analyse the internal structure of the DVjs through topic modeling and derive consequences for distance measurement and simulations.
Quantencomputing in den Digital Humanities: innovativ oder übertrieben? Experimente mit Quanten-basiertem Maschinellen Lernen
Long Paper
von Johanna Barzen und Frank Leymann
Abstract:
With the advent of widely available quantum computers, the question arises why not use the potential of this new technology to address existing or entirely new questions in the Digital Humanities? Especially in the field of quantum-based machine learning, the quantum computers available today can already achieve results that are partially superior to classical solutions. This is, for example, relevant for the Digital Humanities with steadily growing data volumes and accompanying analysis requirements. To explore such potentials of quantum computing, we would like to present QHAna, which allows the comparison of the results of classical and quantum-based machine learning methods.
Wie sich die Bilder ähneln. Vom Zufallsfund zur systematischen Forschung im Bereich der automatisierten Bildähnlichkeitssuche
Long Paper
von Wiebke Helm, Sebastian Schmideler, Chanjong Im, Thomas Mandl, Stefanie Kollmann, Lars Müller
Abstract:
The perception of the transmission of ideas and knowledge through texts is well-known to us. But the image also has a »life of its own« in relation to the text that accompanies and explains it. This has hardly been taken into notice so far. The paper takes up this problem using the example of historical children’s and youth literature and focuses here in particular on the genre of books teaching knowledge. In this way, the contexts of origin of the book illustrations will be shown and their references reconstructed. In consequence, the didactic selection of the image content, which was considered relevant for the education of the child in the 18th and 19th century, can be better understood. The way to a solution is interdisciplinary and combines the library perspective in digitizing and providing the image sources, two heterogeneous information science methods for similarity search and the subject-specific expertise in children’s book research. We will compare and critically reflect on the suitability and scope of the methods.
Experiments in the digital laboratory. What the Computational Humanities can learn about their definition and terminology from the History of Science
Long Paper
von Sarah Lang
Abstract:
The notion of experiment is widely discussed in Digital Humanities, yet a precise terminology is still lacking. This article argues that a narrow set of scenarios in the subfield of the Computational Humanities can implement a definition close to that of the Sciences which transcends the metaphor of playful exploration. It proposes that we should look to those Humanities disciplines which have already integrated experimental methods into their hermeneutic arsenal such as Experimental Archaeology and the Experimental History of Science. These share an emphasis on distinguishing ›experiencing‹ from ›experimenting‹ which I argue should be applied to the Digital Humanities notion of experiment as well.
A Sentiment Analysis Tool Chain for 18th Century Periodicals
Code Experiment
von Philipp Koncar, Bernhard C. Geiger, Christina Glatz, Elisabeth Hobisch, Sanja Sarić, Martina Scholger, Yvonne Völkl, Denis Helic
Abstract:
Sentiment analysis is a common task in natural language processing (NLP) and aims for the automatic and computational identification of emotions, attitudes and opinions expressed in textual data. While Sentiment analysis is typically tailored for and widely used in the context of Web data, the application to literary texts is still challenging due to the lack of methods dedicated to languages other than English and from earlier times. With the work we present here, we not only introduce new sentiment dictionaries for French, Italian and Spanish periodicals of the 18th century, but also build a freely and publicly available tool chain based on Jupyter Notebooks, enabling researchers to apply our dictionary creation process and sentiment analysis methods to their own material and projects. The proposed tool chain comprises two different parts: (i) the optional creation of sentiment dictionaries and (ii) the actual sentiment analysis.
»Embed, embed! There’s knocking at the gate.« Detecting Intertextuality with Embeddings and the Vectorian
Code Experiment
von Bernhard Liebl und Manuel Burghardt
Abstract:
The detection of intertextual references in text corpora is a digital humanities topic that has gained a lot of attention in recent years. While intertextuality – from a literary studies perspective – describes the phenomenon of one text being present in another text, the computational problem at hand is the task of text similarity detection, and more concretely, semantic similarity detection. In this notebook, we introduce the Vectorian as a framework to build queries through word embeddings such as fastText and GloVe. We evaluate the influence of computing document similarity through alignments such as Waterman-Smith-Beyer and two variants of Word Mover’s Distance. We also investigate the performance of state-of-art sentence embeddings like Siamese BERT networks for the task – both as document embeddings and as contextual token embeddings. Overall, we find that Waterman-Smith-Beyer with fastText offers highly competitive performance. The notebook can also be used to upload new data for performing custom search queries.