Search

Bereits seit einigen Jahren werden große Anstrengungen unternommen, um die im deutschen Sprachraum erschienenen Drucke des 16.-18. Jahrhunderts zu erfassen und zu digitalisieren. Deren Volltexttransformation konzeptionell und technisch vorzubereiten, ist das übergeordnete Ziel des DFG-Projekts OCR-D, das sich mit der Weiterentwicklung von Verfahren der Optical Character Recognition befasst. Der Beitrag beschreibt den aktuellen Entwicklungsstand der OCR-D-Software und analysiert deren erste Teststellung in ausgewählten Bibliotheken., For some years, great efforts have been made to record and digitise the printed works of the 16th-18th centuries published in the German-speaking countries. Preparing their full-text transformation conceptually and technically is the overall goal of the DFG-project OCR-D, which is concerned with the further development of optical character recognition methods. The article describes the current state of development of the OCR-D software and analyses its initial testing in selected libraries.

We present the Game Walkthrough Corpus (GWTC), which contains 12,295 unique walkthrough documents covering 6,117 games. For each game walkthrough, we provide frequencies of unigrams and bigrams, treating the walkthrough document as a Bag of Words. In addition, we provide word frequencies at the sentence level. Furthermore, the GWTC contains a number of game-related metadata, including title, publisher, developer, year, and genre. All the language statistics and metadata are stored in separate plain text files and can be referenced through uniform resource names (URN). These URNs can also be used to derive any combination of statistics and metadata. Researchers, for instance, can investigate the most frequent unigrams for games in the “Adventure” genre. This way, the GWTC can be reused for different kinds of research questions on gaming language.

One important and particularly challenging step in the optical character recognition (OCR) of historical documents with complex layouts, such as newspapers, is the separation of text from non-text content (e.g. page borders or illustrations). This step is commonly referred to as page segmentation. While various rule-based algorithms have been proposed, the applicability of Deep Neural Networks (DNNs) for this task recently has gained a lot of attention. In this paper, we perform a systematic evaluation of 11 different published DNN backbone architectures and 9 different tiling and scaling configurations for separating text, tables or table column lines. We also show the influence of the number of labels and the number of training pages on the segmentation quality, which we measure using the Matthews Correlation Coefficient. Our results show that (depending on the task) Inception-ResNet-v2 and EfficientNet backbones work best, vertical tiling is generally preferable to other tiling approaches, and training data that comprises 30 to 40 pages will be sufficient most of the time., Evaluation of deep neural networks for the segmentation of pages of historical newspapers; 21 pages total (incl. references and appendix), 7 figures, 5 tables

In this paper, we discuss the computer-aided processing of handwritten tabular records of historical weather data. The observationes meteorologicae, which are housed by the Regensburg University Library, are one of the oldest collections of weather data in Europe. Starting in 1771, meteorological data was consistently documented in a standardized form over almost 60 years by several writers. The tabular structure, as well as the unconstrained textual layout of comments and the use of historical characters, propose various challenges in layout and text recognition. We present a customized strategy to digitize tabular and handwritten data by combining various state-of-the-art methods for OCR processing to fit the collection. Since the recognition of historical documents still poses major challenges, we provide lessons learned from experimental testing during the first project stages. Our results show that deep learning methods can be used for text recognition and layout detection. However, they are less efficient for the recognition of tabular structures. Furthermore, a tailored approach had to be developed for the historical meteorological characters during the manual creation of ground truth data. The customized system achieved an accuracy rate of 82% for the text recognition of the heterogeneous handwriting and 87% accuracy for layout recognition of the tables.

Der Beitrag beschreibt ein laufendes Projekt zur computergestützten Erschließung und Analyse einer großen Sammlung handschriftlicher Liedblätter mit Volksliedern aus dem deutschsprachigen Raum. Am Beispiel dieses praktischen Projekts werden Chancen und Herausforderungen diskutiert, die der Einsatz von Digital Humanities-Methoden für den Bereich der Musikwissenschaft mit sich bringt. This article presents an ongoing project for the computer-based transcription and analysis of handwritten music scores from a large collection of German folk tunes. Based on this project, I will discuss the challenges and opportunities that arise when using Digital Humanities methods in musicology.

Patterns are a popular means to document design knowledge in different fields of application, including HCI. Accordingly, a large number of different HCI pattern formats have been suggested. However, relatively little is said about how to systematically identify such patterns. In order to make the process of identifying patterns more transparent and comprehensible, I apply a usability inspection method to generate data that can be used as input for the systematic creation of usability patterns. This article describes the basic idea of identifying usability patterns through a series of heuristic walkthroughs and illustrates the approach by means of a case study in the field of “linguistic annotation tools”.

In this article we describe a usability evaluation of eight desktop search engines (DSEs). We used the heuristic walkthrough method to gather usability problems as well as individual strengths and weaknesses of the tested search engines. The results of the evaluation are integrated into a set of 30 design guidelines for user-friendly DSEs.

Walk-up-and-use-systems such as vending and self-service machines request special attention concerning an easy to use and self-explanatory user interface. In this paper we present a set of design guidelines for coffee vending machines based on the results of an expert-based usability evaluation of thirteen different models.

We present a web-based tool for evaluating the information architecture of a website. The tool allows the use of crowdsourcing platforms like Amazon's MTurk as a means for recruiting test persons, and to conduct asynchronous remote navigation stress tests (cf. Instone 2000). We also report on an evaluation study which compares our tool-based crowdsourced approach to a more traditional laboratory test setting. Results of this comparison indicate that although there are interesting differences between the two testing approaches, both lead to similar test results.

Purpose — This chapter illustrates and explains the ambiguity and vagueness of the term social search and aims at describing and classifying the heterogeneous landscape of social search implementations on the WWW. Methodology/approach — We have looked at different definitions as well as the context of social search by carrying out an extensive literature review, and tried to unify and enhance existing ideas and concepts. Our definition of social search is illustrated by a general review of existing social search engines, which are analyzed and described by their specific features and social aspects. Findings — The chapter presents a discussion of social search as well as a comparison of existing social search engines. Social implications — The definition of social search and the comparison of social search engines summarize the many ways people can search the web together and allow for an assessment of future developments in this area. Originality/value of paper — Although different attempts to define social search have been made in the past, we present an argumentation that unifies some existing definitions and which is different from other interpretations of the social search concept. We present an overview and a comparison of the different genres of social search engines.

Books printed before 1800 present major problems for OCR. One of the main obstacles is the lack of diversity of historical fonts in training data. The OCR-D project, consisting of book historians and computer scientists, aims to address this deficiency by focussing on three major issues. Our first target was to create a tool that identifies font groups automatically in images of historical documents. We concentrated on Gothic font groups that were commonly used in German texts printed in the 15th and 16th century: the well-known Fraktur and the lesser known Bastarda, Rotunda, Textura und Schwabacher. The tool was trained with 35,000 images and reaches an accuracy level of 98%. It can not only differentiate between the above-mentioned font groups but also Hebrew, Greek, Antiqua and Italic. It can also identify woodcut images and irrelevant data (book covers, empty pages, etc.). In a second step, we created an online training infrastructure (okralact), which allows for the use of various open source OCR engines such as Tesseract, OCRopus, Kraken and Calamari. At the same time, it facilitates training for specific models of font groups. The high accuracy of the recognition tool paves the way for the unprecedented opportunity to differentiate between the fonts used by individual printers. With more training data and further adjustments, the tool could help to fill a major gap in historical research.