The medieval legendary "Der Heiligen Leben, Redaktion" exists in two variants. As part of an ongoing digital edition project, the difference phenomena between these variants are manually annotated and categorized. We present an interactive, comparative visualization approach for the exploration of these annotation data with the goal of analyzing the nature of text variance in this work. The approach combines a set of coordinated close and distant reading views in an overview+detail design and provides means for category-filtering and detailed inspection of the data. We conducted a task and requirement analysis in cooperation of the involved medievalists and visualization researchers and sketch two scenarios of usage.

Today, libraries provide digitized collections of historical newspapers, which researchers in the humanities seek to analyze. An important objective of this work is to enable researchers to overview and analyze the textual, temporal and geographical dissemination of an event expressed in document corpora of interest. For this, we propose LilyPads, which permits researchers to analyze such corpora using a novel, map-inset-based approach. In contrast to previous work, LilyPads is centered around one main view, which integrates key aspects of the visualized data, thereby facilitating an explorative approach to finding relationships in data. From LilyPads' overview, researchers can select subsets of data as well as individual documents interactively, which supports detailed analysis of the corpus, combining close and distant reading methods. We show the applicability of LilyPads by demonstrating its use in a real-world analysis scenario.

Interactive visual analysis of documents relies critically on the ability of machines to process and analyze texts. Important techniques for text processing include text summarization, classification, or translation. Many of these approaches are based on part-of-speech tagging, a core natural language processing technique. Part-of-speech taggers are typically trained on collections of modern newspaper, magazine, or journal articles. They are known to have high accuracy and robustness when applied to contemporary newspaper style texts. However, the performance of these taggers deteriorates quickly when applying them to more domain specific writings, such as older or even historical documents. Large training sets tend to be scarce for these types of texts due to the limited availability of source material and costly digitization and annotation procedures. In this paper, we present an interactive visualization approach that facilitates analysts in determining part-of-speech tagging errors by comparing several standard part-of-speech tagger results graphically. It allows users to explore, compare, evaluate, and adapt the results through interactive feedback in order to obtain a new model, which can then be applied to similar types of texts. A use case shows successful applications of the approach and demonstrates its benefits and limitations. In addition, we provide insights generated through expert feedback and discuss the effectiveness of our approach.

Magic lens based focus+context techniques are powerful means for exploring document spatializations. Typically, they only offer additional summarized or abstracted views on focused documents. As a consequence, users might miss important information that is either not shown in aggregated form or that never happens to get focused. In this work, we present the design process and user study results for improving a magic lens based document exploration approach with exemplary visual quality cues to guide users in steering the exploration and support them in interpreting the summarization results. We contribute a thorough analysis of potential sources of information loss involved in these techniques, which include the visual spatialization of text documents, user-steered exploration, and the visual summarization. With lessons learned from previous research, we highlight the various ways those information losses could hamper the exploration. Furthermore, we formally define measures for the aforementioned different types of information losses and bias. Finally, we present the visual cues to depict these quality measures that are seamlessly integrated into the exploration approach. These visual cues guide users during the exploration and reduce the risk of misinterpretation and accelerate insight generation. We conclude with the results of a controlled user study and discuss the benefits and challenges of integrating quality guidance in exploration techniques.

The analysis of a novel's plot and characters are challenging and time-consuming tasks in literary criticism. Typically, humanities scholars want to describe and compare characters' personality traits, their roles, their relationships, and the evolution of these aspects over the course of a novel. Nowadays, due to the digitization of literature, humanities scholars can be supported in these endeavors with computational methods. In this paper, we present an approach that offers several means to analyze the plot and characters of a novel visually. Analysts can easily switch between an adjacency matrix and a node-link representation, which provide an overview of the characters and the relationships between them. Both views enable analysts to select different text ranges of the novel for studying the commonalities and differences of the character constellations within these ranges. We offer interactive visual representations to help investigate the relationships between the characters in more detail. Additionally, we link the visual representations with the novels' texts to support the inspection and verification of previously generated ideas and hypotheses. To demonstrate the benefits and limitations of our approach, we present two usage scenarios. The first one is based on a fictitious analysis and the second one discusses applications that were carried out during joint workshops with humanities scholars. Finally, we present and discuss the insights gained by an expert study and the design decisions of our approach.

The increasing availability of digital multimedia content has led to the need of new approaches for the analysis of large databases containing video and associated data, for example, subtitles. Visualization provides valuable insights of such dataset, complementing approaches solely based on techniques for knowledge discovery in databases and information retrieval. Hence, visual analytics, combining automatic processing with interactive data visualization, has proven to be an effective means to explore and interpret such data. The analysis of news corpora represents a typical task for such a scenario. Domain experts such as journalists and social science scholars require an overview of important topics, the temporal coherence of events, and they should be able to compare different topics. We present a visual analytics approach that aims to support these tasks with automatic video preprocessing, topic extraction, clustering, and dimensionality reduction. Coordinated linked views support the flexible inspection of the dataset and the processed results. We further discuss the application of our approach in a usage scenario, inspecting the dataset of a daily news broadcast of the year 2015.

Interactive text visualization can help users explore and gain insights into complex and often large document sets. One popular visualization strategy to represent such collections is to depict each document as a glyph in 2D space. These spaces have proven effective, especially when combined with interactive exploration methods. However, current exploratory approaches are largely limited to single areas of a 2D spatialization, lacking support for important comparative exploration and analysis tasks. In this paper, we extend a flexible focus+context exploration technique to tackle this challenge. In particular, based on practical tasks from the digital humanities, we focus on exploring and investigating relationships between entities in large document collections. Our approach uses natural language processing to extract characters and places, including information about their relationships. We then use linked views to facilitate visual analysis of extracted information artifacts. Based on two usage scenarios, we demonstrate successful applications of the approach and discuss its benefits and limitations.

In recent years, interactive visualization to analyze text documents has gained an impressive momentum. This is not surprising considering the fast increase of electronically available textual documents of various kinds. These include, for example, patents, scholarly documents, social media messages, and many other sources that contain valuable knowledge and insights for many stakeholders. Interactive text visualization turned out to be an important means for exploring and gaining insights into complex and often large document collections. An established visualization strategy to represent such collections is using projection-based techniques that visualize documents as glyphs in a 2D view aiming to reflect the semantic similarity of documents by the proximity of their placement. Static labels have been suggested to characterize the overall topics contained in the projected data to improve the effectiveness of such visualization techniques. Other approaches employ magic lenses that enable users to explore the 2D spatialization freely on various granularity levels. In this work, we propose a visual exploration approach that combines cluster-based labeling of projected documents with an interaction concept for magic lens techniques. We offer a set of novel interactive features to support a smooth transition between static labels and the magic lens approach while exploiting the different levels of visual abstraction of both techniques without introducing additional clutter through overdraw. Finally, we provide insights gained from a preliminary user study and present the benefits of our approach.

Word Clouds have gained an impressive momentum for summarizing text documents in the last years. They visually communicate in a clear and descriptive way the most frequent words of a text. However, there are only very few word cloud visualizations that support a contrastive analysis of multiple documents. The available approaches provide comparable overviews of the documents, but have shortcomings regarding the layout, readability, and use of white space. To tackle these challenges, we propose MultiCloud, an approach to visualize multiple documents within a single word cloud in a comprehensible and visually appealing way. MultiCloud comprises several parameters and visual representations that enable users to alter the word cloud visualization in different aspects. Users can set parameters to optimize the usage of available space to get a visual representation that provides an easy visual association of words with the different documents. We evaluated MultiCloud with visualization researchers and a group of domain experts comprising five humanities scholars.

The annotation of video material plays an important role in many Digital Humanities research fields including arts, political sciences, and cultural and historical studies. The annotations are typically assigned manually and convey rich semantics in accordance with the respective research question. In this work, we present the concept of a visual analytics approach that enables researchers to annotate multiple video sources in parallel. It combines methods from the fields of natural language processing and computer vision to support the manual annotation process with automatically extracted lowlevel characteristics. The benefits of our approach are twofold. With the extracted annotations and their visual mapping onto a suitable overview visualization, we support scholars in finding the relevant sections for their high-level annotations on the one hand, and on the other hand, we offer an environment that lets them compare and analyze such annotations in several videos at once. Our concept can be flexibly extended with additional processing methods to simplify annotation tasks further.

The study of novels and the analysis of their plot, characters and other information entities are complex and time-consuming tasks in literary science. The digitization of literature and the proliferation of electronic books provide new opportunities to support these tasks with visual abstractions. Methods from the field of computational linguistics can be used to automatically extract entities and their relations from digitized novels. However, these methods have known limitations, especially when applied to narrative text that does often not follow a common schema but can have various forms. Visualizations can address the limitations by providing visual clues to show the uncertainty of the extracted information, so that literary scholars get a better idea of the accuracy of the methods. In addition, interaction can be used to let users control and adapt the extraction and visualization methods according to their needs. This paper presents ViTA, a web-based approach that combines automatic analysis methods with effective visualization techniques. Different views on the extracted entities are provided and relations between them across the plot are indicated. Two usage scenarios show successful applications of the approach and demonstrate its benefits and limitations. Furthermore, the paper discusses how uncertainty might be represented in the different views and how users can be enabled to adapt the automatic methods.

The Academy of Sciences and Literature in Mainz provides online access to the Regesta Imperii a very extensive historical dataset based on documentary sources of German-Roman kings. About 125,000 regestae of emperors and popes are searchable and viewable in this online portal. The current user interface offers direct access to single documents through different form-based search facilities, as well as through a catalogue that directly reflects the structure of the regestae volumes as they have been created in this long-term project. In order to further improve access to this large data volume, we suggest an additional approach based on coordinated views. The usage of coordinated view approaches is very common in many domains (Stasko et al, 2008, Vuillemot et al, 2009, Koch et al, 2011). However, there is no publicly system available, which would provide a suitable access to this historical dataset. The motivation for this new approach is twofold. On the one hand, we improve the support for search and exploration tasks in this historical data set that are based on imprecise information needs or on a less deep understanding of the available information. In practice, such imprecise queries can quickly lead to an overwhelmingly large number of search results. Allowing users to create and refine queries in a visual way, while offering immediate feedback on the number of entries requested, can help to cope with underspecified queries and help refining them iteratively (Jänicke et al, 2012). On the other hand, we offer a powerful means for visually analyzing the available information and understanding complex relationships by providing different linked perspectives on subsets of the collections. These perspectives include views on historic persons and entities as well as temporal and spatial information contained in the regestae. A usage scenario shows successful application of the approach.

This paper presents an approach to extract co-occurrence networks from literary texts. It is a deliberate decision not to aim for a fully automatic pipeline, as the literary research questions need to guide both the definition of the nature of the things that co-occur as well as how to decide co-occurrence. We showcase the approach on a Middle High German romance, Parzival. Manual inspection and discussion shows the huge impact various choices have.

Evaluation has become a fundamental part of visualization research and researchers have employed many approaches from the field of human-computer interaction like measures of task performance, thinking aloud protocols, and analysis of interaction logs. Recently, eye tracking has also become popular to analyze visual strategies of users in this context. This has added another modality and more data, which requires special visualization techniques to analyze this data. However, only few approaches exist that aim at an integrated analysis of multiple concurrent evaluation procedures. The variety, complexity, and sheer amount of such coupled multi-source data streams require a visual analytics approach. Our approach provides a highly interactive visualization environment to display and analyze thinking aloud, interaction, and eye movement data in close relation. Automatic pattern finding algorithms allow an efficient exploratory search and support the reasoning process to derive common eye-interaction-thinking patterns between participants. In addition, our tool equips researchers with mechanisms for searching and verifying expected usage patterns. We apply our approach to a user study involving a visual analytics application and we discuss insights gained from this joint analysis. We anticipate our approach to be applicable to other combinations of evaluation techniques and a broad class of visualization applications.

The creation of interactive visualization to analyze text documents has gained an impressive momentum in recent years. This is not surprising in the light of massive and still increasing amounts of available digitized texts. Websites, social media, news wire, and digital libraries are just few examples of the diverse text sources whose visual analysis and exploration offers new opportunities to effectively mine and manage the information and knowledge hidden within them. A popular visualization method for large text collections is to represent each document by a glyph in 2D space. These landscapes can be the result of optimizing pairwise distances in 2D to represent document similarities, or they are provided directly as meta data, such as geo-locations. For well-defined information needs, suitable interaction methods are available for these spatializations. However, free exploration and navigation on a level of abstraction between a labeled document spatialization and reading single documents is largely unsupported. As a result, vital foraging steps for task-tailored actions, such as selecting subgroups of documents for detailed inspection, or subsequent sense-making steps are hampered. To fill in this gap, we propose DocuCompass, a focus+context approach based on the lens metaphor. It comprises multiple methods to characterize local groups of documents, and to efficiently guide exploration based on users' requirements. DocuCompass thus allows for effective interactive exploration of document landscapes without disrupting the mental map of users by changing the layout itself. We discuss the suitability of multiple navigation and characterization methods for different spatializations and texts. Finally, we provide insights generated through user feedback and discuss the effectiveness of our approach.

In information visualization, evaluation plays a crucial role during the development of a new visualization technique. In recent years, eye tracking has become one means to analyze how users perceive and understand a new visualization system. Since most visualizations are highly interactive nowadays, a study should take interaction, in terms of user-input, into account as well. In addition, think aloud data gives insights into cognitive processes of participants using a visualization system. Typically, researchers evaluate these data sources separately. However, we think it is beneficial to correlate eye tracking, interaction, and think aloud data for deeper analyses. In this paper, we present challenges and possible solutions in triangulating user behavior using multiple evaluation data sources. We describe how the data is collected, synchronized, and analyzed using a string-based and a visualization-based approach founded on experiences from our current research. We suggest methods how to tackle these issues and discuss benefits and disadvantages. Thus, the contribution of our work is twofold. On the one hand, we present our approach and the experiences we gained during our research. On the other hand, we investigate additional methods that can be used to analyze this multi-source data.

The analysis of inherent structures of movies plays an important role in studying stylistic devices and specific, content-related questions. Examples are the analysis of personal constellations in movie scenes, dialogue-based content analysis, or the investigation of image-based features. We provide a visual analytics approach that supports the analytical reasoning process to derive higher level insights about the content on a semantic level. Combining automatic methods for semantic scene analysis based on script and subtitle text, we perform a low-level analysis of the data automatically. Our approach features an interactive visualization that allows a multilayer interpretation of descriptive features to characterize movie content. For semantic analysis, we extract scene information from movie scripts and match them with the corresponding subtitles. With text- and image-based query techniques, we facilitate an interactive comparison of different movie scenes on an image and on a semantic level. We demonstrate how our approach can be applied for content analysis on a popular Hollywood movie.