Juliane Stiller

Automatic Enrichments with Controlled Vocabularies in Europeana: Challenges and Consequences

Automatic enrichment of metadata is one option for digital libraries to add multilingual terms to their resources. Adding links to external vocabularies further contextualizes the metadata in a linked data environment. This paper reports on a case study using the digital library Europeana, which implements this type of automatic contextualization and enrichment strategy. A group of content and metadata experts analyzed seven datasets in Europeana and evaluated the successes and challenges of the implemented enrichment strategy. For each of the three enrichment process stages (analysis, linking and augmentation), common problems and errors were identified. The paper further categorizes the enrichment challenges by original cause and describes consequences of the enrichment errors for the user interactions within digital libraries. Results serve as best practice examples for analyzing enrichment challenges and provide insights and recommendations for other large digital libraries implementing similar enrichment or contextualization strategies.

Poisonous India or the Importance of a Semantic and Multilingual Enrichment Strategy

Cultural heritage information systems offer access to objects coming from museums, archives and libraries. To enhance retrieval performance and access across languages, metadata is enriched with controlled vocabularies or other datasets with structured information. During this process many pitfalls occur which lead to wrong or poor enrichments thus decreasing the user experience. Taking the use case of Europeana, this paper investigates the extent of enrichment flaws and their causes. A categorization of these deficiencies is proposed as well as a strategy to avoid common enrichment mistakes.

Exploring Comparative Evaluation of Semantic Enrichment Tools for Cultural Heritage Metadata

Semantic enrichment of metadata is an important and difficult problem for digital heritage efforts such as Europeana. This paper gives motivations and presents the work of a recently completed Task Force that addressed the topic of evaluation of semantic enrichment. We especially report on the design and the results of a comparative evaluation experiment, where we have assessed the enrichments of seven tools (or configurations thereof) on a sample benchmark dataset from Europeana.

Which log for which information? gathering multilingual data from different log file types

In this paper, a comparative analysis of different log file types and their potential for gathering information about user behavior in a multilingual information system is presented. It starts with a discussion of potential questions to be answered in order to form an appropriate view of user needs and requirements in a multilingual information environment and the possibilities of gaining this information from log files. Based on actual examples from the Europeana portal, we compare and contrast different types of log files and the information gleaned from them. We then present the Europeana Clickstream Logger, which logs and gathers extended information on user behavior, and show first examples of the data collection possibilities.

Searching for inspiration: user needs and search architecture in europeana collections

Drawing upon research and current development work at Europeana, this paper discusses search functionality in the Cultural Heritage sector, focusing in particular on the question of ‘inspiration‐oriented’ search, in which users seek out previously‐unknown items to serve as creative stimulus. Inspiration‐oriented search is identified as a variant of the more widely‐studied problem of serendipitous retrieval, and defined as an information‐seeking behavior in which users consciously search for items that are related to known items in ways that are recognizable once seen, but that are nevertheless unpredictable at search‐time. Various strategies for the maximization of both the recognisability and unpredictability of related items are then described, including user‐interface and user‐experience changes and the reconceptualization of datastores as knowledge graphs. Directions for further research are then outlined – including, most importantly, possible metrics for inspiration‐oriented search and their potential for use in machine‐learning ranking frameworks.

Multilingual Access to Digital Libraries: The Europeana Use Case

The article determines multilingual access features in digital libraries with a special focus on cultural heritage digital libraries. An analysis of existing information systems in the GLAM-domain (galleries, libraries, archives and museums) was conducted to establish and collect solutions for searching, browsing and interacting with multilingual content. In particular, Europeana, the European digital library, archive and museum for cultural heritage was studied with a focus on multilingual interactions. Challenges and recommendations for the implementation of multilingual access features are presented and discussed. Der Artikel fasst Komponenten für einen mehrsprachigen Zugang in digitalen Bibliotheken zusammen. Dabei wird der Fokus auf Bibliotheken für das digitale Kulturerbe gelegt. Eine Analyse aktueller (existierender) Informationssysteme im sogenannten GLAM-Bereich (Galerien, Bibliotheken, Archive, Museen) beschreibt angewandte Lösungen für die Recherche (Suchen und Blättern) von und die Interaktion mit mehrsprachigen Inhalten. Europeana, die europäische digitale Bibliothek für Kulturerbe, wird als Fallbeispiel hervorgehoben und es werden beispielhaft Interaktionsszenarios für die mehrsprachige Recherche vorgestellt. Die Herausforderungen in der Implementierung von Komponenten für den mehrsprachigen Informationszugang sowie Empfehlungen für den verbesserten Einsatz werden vorgestellt und diskutiert. L’article détermine les caractéristiques d’accès multilingues dans les bibliothèques numériques et en particulier dans les bibliothèques numériques dédiées au patrimoine culturel. Une analyse de systèmes d’information existants dans le domaine GLAM (galeries, bibliothèques, archives, musées) a été menée à bien pour fixer et percevoir des solutions pour la recherche et la navigation de contenu multilingue. En particulier, Europeana, la bibliothèque numérique européenne dédiée au patrimoine culturel, a été étudiée avec un accent sur les interactions de recherche multilingue. Les défis de la mise en oeuvre des fonctions d’accès multilingues et des recommandations sur la manière de les surmonter sont ici présentés et discutés.

A Framework for Classifying Interactions in Cultural Heritage Information Systems

With the mass digitization of cultural heritage and the increase of people accessing the digitized memory objects, it becomes crucial to develop meaningful interaction patterns in cultural heritage information systems. This explorative study is based on an investigation of 50 websites from the cultural heritage domain. It derives a framework for classifying user interactions with digital cultural heritage. The framework has two dimensions; the first one is a classification of the interactions and the second one describes their degree of complexity. The strength of this framework is the ability to compare complexity, scope and purpose of interactions across different websites while offering a meaningful vocabulary for discussing different interaction features.

Is tagging multilingual?: a case study with BibSonomy

This paper investigates the occurrence of tags in different languages in a collaborative bookmarking and publication sharing service - BibSonomy. Social tags assigned to URLs in multiple languages and users tagging these URLs multilingually are the main focus of this study. The results show that multilingual tags occur for the same URL and that users tag in different languages. Furthermore, the results give indications that the language of the content of a URL does not imply that its tags are in the same language.

Leveraging user interaction and collaboration for improving multilingual information access in digital libraries

The goal of interactive cross-lingual information retrieval systems is to support users in formulating effective queries and selecting the documents which satisfy their information needs regardless of the language of these documents. This dissertation aims at harnessing user-system interaction, extracting the added value and integrating it back into the system to improve cross-lingual information retrieval for successive users. To achieve this, user input at different interaction points will be evaluated. This will, among others, include interaction during user-assisted query translations, implicit and explicit relevance feedback and social tags. To leverage this input, explorative studies need to be conducted to determine beneficial user input and the methods of extracting it.

Assessing Digital Skills of Refugee Migrants During Job Orientation in Germany

This paper examines the digital skill level of refugee migrants in Germany while pursuing a job, a training position, or following an educational path on the Internet. For that, we conducted a lab experiment designing tasks with varying difficulty to position the digital competencies of refugee migrants on the digital skill scale. Problems with operational and formal skills were observed whereas fact-based information seeking was often successfully completed. The most complex tasks could not be completed by any participant. The study contributes to a better understanding of the varying degrees of digital skills of refugee migrants. Results can be used to design targeted courses and curricula that address digital deficits. Further training in this area will enable refugee migrants to benefit from the many opportunities that arise through the Internet and its services, improving their chances for labor market integration.