Tschinkel Gerwald
2020
One classic issue associated with being a researcher nowadays is the multitude and magnitude of search results for a given topic. Recommender systems can help to fix this problem by directing users to the resources most relevant to their specific research focus. However, sets of automatically generated recommendations are likely to contain irrelevant resources, making user interfaces that provide effective filtering mechanisms necessary.This problem is exacerbated when users resume a previously interrupted research task, or when different users attempt to tackle one extensive list of results, as confusion as to what resources should be consulted can be overwhelming.The presented recommendation dashboard uses micro-visualisations to display the state of multiple filters in a data type-specific manner. This paper describes the design and geometry of micro-visualisations and presents results from an evaluation of their readability and memorability in the context of exploring recommendation results. Based on that, this paper also proposes applying micro-visualisations for extending the use of a desktop-based dashboard to the needs of small-screen, mobile multi-touch devices, such as smartphones. A small-scale heuristic evaluation was conducted using a first prototype implementation.
Tschinkel Gerwald, Sabol Vedran
2017
When using classical search engines, researchers are often confronted with a number of results far beyond what they can realistically manage to read; when this happens, recommender systems can help, by pointing users to the most valuable sources of information. In the course of a long-term research project, research into one area can extend over several days, weeks, or even months. Interruptions are unavoidable, and, when multiple team members have to discuss the status of a project, it’s important to be able to communicate the current research status easily and accurately. Multiple type-specific interactive views can help users identify the results most relevant to their focus of interest. Our recommendation dashboard uses micro-filter visualizations intended to improve the experience of working with multiple active filters, allowing researchers to maintain an overview of their progress. Within this paper, we carry out an evaluation of whether micro-visualizations help to increase the memorability and readability of active filters in comparison to textual filters. Five tasks, quantitative and qualitative questions, and the separate view on the different visualisation types enabled us to gain insights on how micro-visualisations behave and will be discussed throughout the paper.
Tschinkel Gerwald, Hasitschka Peter, Sabol Vedran, Hafner R
2016
Faceted search is a well known and broadly imple- mented paradigm for filtering information with various types of structured information. In this paper we introduce a multiple-view faceted interface, consisting of one main visualisation for exploring the data and multiple minia- turised visualisations showing the filters. The Recommen- dation Dashboard tool provides several interactive visual- isations for analysing recommender results along various faceted dimensions specific to cultural heritage and scien- tific content. As our aim is to reduce the user load and opti- mise the use of screen area, we permit only one main visu- alisation to be visible at a time, and introduce the concept of micro-visualisations – small, simplified views conveying only the necessary information – to provide natural, easy to understand representation of the the active filter set.
Tschinkel Gerwald, di Sciascio Maria Cecilia, Mutlu Belgin, Sabol Vedran
2015
Recommender systems are becoming common tools supportingautomatic, context-based retrieval of resources.When the number of retrieved resources grows large visualtools are required that leverage the capacity of humanvision to analyse large amounts of information. Weintroduce a Web-based visual tool for exploring and organisingrecommendations retrieved from multiple sourcesalong dimensions relevant to cultural heritage and educationalcontext. Our tool provides several views supportingfiltering in the result set and integrates a bookmarkingsystem for organising relevant resources into topic collections.Building upon these features we envision a systemwhich derives user’s interests from performed actions anduses this information to support the recommendation process.We also report on results of the performed usabilityevaluation and derive directions for further development.
Veas Eduardo Enrique, Mutlu Belgin, di Sciascio Maria Cecilia, Tschinkel Gerwald, Sabol Vedran
2015
Supporting individuals who lack experience or competence to evaluate an overwhelming amout of informationsuch as from cultural, scientific and educational content makes recommender system invaluable to cope withthe information overload problem. However, even recommended information scales up and users still needto consider large number of items. Visualization takes a foreground role, letting the user explore possiblyinteresting results. It leverages the high bandwidth of the human visual system to convey massive amounts ofinformation. This paper argues the need to automate the creation of visualizations for unstructured data adaptingit to the user’s preferences. We describe a prototype solution, taking a radical approach considering bothgrounded visual perception guidelines and personalized recommendations to suggest the proper visualization.
Tschinkel Gerwald, Veas Eduardo Enrique, Mutlu Belgin, Sabol Vedran
2014
Providing easy to use methods for visual analysis of LinkedData is often hindered by the complexity of semantic technologies. Onthe other hand, semantic information inherent to Linked Data providesopportunities to support the user in interactively analysing the data. Thispaper provides a demonstration of an interactive, Web-based visualisa-tion tool, the “Vis Wizard”, which makes use of semantics to simplify theprocess of setting up visualisations, transforming the data and, most im-portantly, interactively analysing multiple datasets using brushing andlinking method
Mutlu Belgin, Tschinkel Gerwald, Veas Eduardo Enrique, Sabol Vedran, Stegmaier Florian, Granitzer Michael
2014
Research papers are published in various digital libraries, which deploy their own meta-models and tech-nologies to manage, query, and analyze scientific facts therein. Commonly they only consider the meta-dataprovided with each article, but not the contents. Hence, reaching into the contents of publications is inherentlya tedious task. On top of that, scientific data within publications are hardcoded in a fixed format (e.g. tables).So, even if one manages to get a glimpse of the data published in digital libraries, it is close to impossibleto carry out any analysis on them other than what was intended by the authors. More effective querying andanalysis methods are required to better understand scientific facts. In this paper, we present the web-basedCODE Visualisation Wizard, which provides visual analysis of scientific facts with emphasis on automatingthe visualisation process, and present an experiment of its application. We also present the entire analyticalprocess and the corresponding tool chain, including components for extraction of scientific data from publica-tions, an easy to use user interface for querying RDF knowledge bases, and a tool for semantic annotation ofscientific data set