24 November 2018. Tagung der deutscher ingenieurinnenbund e.V. Hannover (Invited talk)
Title: Urbane Mobilität – ein Einblick in die Datenanalyse.
Die zunehmende Verfügbarkeit von Daten, wie Verkehrsinformationen oder Webdaten, birgt ein großes Potenzial für ein besseres Verständnis der urbanen Mobilität und effizientere Gestaltung innovativer Mobilitätsdienstleistungen. Die Beantwortung komplexer Fragestellungen wird aber erst durch die Verschränkung und Analyse von vielen heterogenen Datenquellen ermöglicht. Der Vortrag bietet aktuelle Einblicke in das Forschungsprojekt „Data4UrbanMobility – Datenbasierte Mobilitätsdienstleistungen für die Stadt der Zukunft“. In diesem Projekt werden Werkzeuge entwickelt, die einen ereignisbasierten Überblick über Mobilitätsinformationen liefern um effiziente Planung, Entwicklung, Durchführung und Nutzung von innovativen Mobilitätsdienstleistungen, insbesondere in der Region Hannover, zu ermöglichen.
10 October 2018. International Semantic Web Conference (ISWC 2018), BlueSky Track.
Title: Can Children Teach AI? Towards Expressive Human-AI Dialogs. [pdf]
AI-empowered dialog systems become increasingly widespread. Whereas this adoption facilitates AI-based systems to engage in end user dialogs on an unprecedented scale, the ability of AI to learn from their human dialog partners is still substantially limited. In this presentation I briefly discuss the opportunities, challenges and risks on the interface of knowledge acquisition and human computer interaction towards more expressive dialogs between end users and AI.
4 June 2018. 4th International Workshop on Social Media World Sensors – Heraklion, Crete, Greece in conjunction with The 15th European Semantic Web Conference. (Invited talk)
Title: Towards Cross-Lingual Event-Centric Information Spaces.
The amount of multilingual information regarding contemporary and historical events of global importance constantly grows on the web, in the news sources and within social media. Efficient collection and effective analytics of increasingly available large-scale event-centric multilingual information is crucial for a variety of real-world applications in the fields of Semantic Web, NLP and Digital Humanities. Cross-Lingual Event-Centric Information Spaces (CL-ECIS) aim to provide an integrated view on such information in particular domains. This paper briefly discusses selected challenges and recent works on event-centric multilingual data collection, semantic representation, analytics and novel applications that constitute initial steps towards building these spaces.
8 March 2018. Workshop “Zukunft des Forschungsdatenmanagements für Ingenieurinnen und Ingenieure“, Technische Universität Darmstadt. (Impulsvortrag)
Titel: Forschungsdatenmanagement: die Web Science Perspektive [slides]
14 February 2018. Humanistisches Forum Garbsen, Hannover
Titel: Selbstlernende Systeme in der Künstlichen Intelligenz. Verstehen wir uns?
Die geheime Sprache: Ein Beispiel aus der Forschung verständlich erklärt. [slides]
Intelligente Informationssysteme beweisen sich zunehmend als praktische Helfer bei vielen alltäglichen Aufgaben. Gleichzeitig wird es, insbesondere in den Medien, kontrovers über die Möglichkeiten, Grenzen und potentielle Gefahren diese Systeme diskutiert. In dem Vortrag werden diese Zusammenhänge anhand von einem aktuellen Beispiel aus dem Bereich maschinelles Lernen für Dialogsysteme verdeutlicht.
21 August – 25 August 2017. 3rd KEYSTONE Training School, TU Wien, Austria (Tutorial & Hands on)
The amount of unstructured information available on the Web is ever growing. Information Extraction enables to automatically identify information nuggets such as named entities, time expressions, relations and events in text and interlink these information nuggets with structured background knowledge. Extracted information can then be used in many application domains, e.g. to categorize and cluster text, enable faceted exploration, populate knowledge bases, and correlate extracted data with other sources. In this introductory tutorial, we provide an overview of the basic blocks for Information Extraction, including methods for named entity extraction and linking, temporal extraction, relation extraction, and open Information Extraction.
Title: Data4UrbanMobility: Data-Driven Mobility Services for Smart Cities
Abstract: Cities of the future have a growing demand in intelligent mobility services and infrastructure to support better mobility and enhance quality of life in urban areas. The goal of the Data4UrbanMobility project is to support this demand through aggregation and analysis of mobility-related data from heterogeneous sources, in particular data about events, public transportation infrastructure and usage, oating car data, as well as behavior and perception of users. While data is spread across heterogeneous institutional repositories, Web platforms and in particular the social Web, semantic technologies and machine learning methods will be exploited to enable the extraction and analysis of data.
8 December 2016. University of Southampton, UK (Tutorial) [slides]
Title: Data Visualisation: Data visualisation tools & Twitter data
The aim of the tutorial is to demonstrate the applications of visualisation techniques for data analysis at the example of Twitter data and to develop practical skills to create such visualisations.
Title: Introduction to Information Extraction
Abstract: Information Extraction enables to automatically identify information nuggets such as named entities, time expressions, relations and events in text and interlink these information nuggets with structured background knowledge. Extracted information can then be used in many application domains, e.g. to categorize and cluster text, enable faceted exploration, populate knowledge bases, and correlate extracted data with other sources. In this introductory tutorial, we provide an overview of the basic blocks for Information Extraction, including methods for named entity extraction and linking, temporal extraction, relation extraction, and open Information Extraction.
17 – 21 July, 2016. ACM SIGIR 2016, Pisa, Italy. (Demo paper)
Title: Analysing Temporal Evolution of Interlingual Wikipedia Article Pairs.
Simon Gottschalk, Elena Demidova.
Abstract: Wikipedia articles representing an entity or a topic in difffferent language editions evolve independently within the scope of the language-specific user communities. This can lead to different points of views reflected in the articles, as well as complementary and inconsistent information. An analysis of how the information is propagated across the Wikipedia language editions can provide important insights in the article evolution along the temporal and cultural dimensions and support quality control. To facilitate such analysis, we present MultiWiki – a novel web-based user interface that provides an overview of the similarities and differences across the article pairs originating from different language editions on a timeline. MultiWiki enables users to observe the changes in the interlingual article similarity over time and to perform a detailed visual comparison of the article snapshots at a particular time point.
Title: Interactive keyword-based access to large scale structured datasets.
Abstract: The data available on the Web, in large scale Web archives, in digital libraries and in open datasets is constantly growing and changing its appearance. The structural and linguistic heterogeneity of this data, as well as its large scale, substantially restrict its accessibility to the end users. In order to supply users with relevant and fresh information on demand, effective and efficient methods are essential that are able to cope with unknown data structures and large scale data. In this tutorial, we explore methods tackling these challenges with the focus on interactive retrieval techniques for structured data that do not require a-priori schema knowledge.
24 May 2016. ACM WebSci, Hannover, Germany. (Short paper)
Title: Analyzing Web archives through Topic and Event Focused Sub-Collections.
Gerhard Gossen, Elena Demidova and Thomas Risse.
Abstract: Web archives capture the history of the Web and are therefore an important source to study how societal developments have been reflected on the Web. However, the large size of Web archives and their temporal nature pose many challenges to researchers interested in working with these collections. In this work, we describe the challenges of working with Web archives and propose the research methodology of extracting and studying sub-collections of the archive focused on specific topics and events. We discuss the opportunities and challenges of this approach and suggest a framework for creating sub-collections.
Title: Efficient extraction of event-centric sub-collections from the Web and large scale Web archives.
Abstract: The Web and Web archives are invaluable sources to follow the traces of recent and past events, in particular for researchers in the Digital Humanities, journalists and historians. On the one hand, the large size of data and their distributed nature makes their analysis daunting, especially for non-computer scientists. On the other hand, most research questions only require a smaller relevant subset of the Web or a Web archive such as the snapshots of Web pages describing one particular event or topic. For example, these sub-collections can reflect the ongoing refugee crisis in Europe, the Fukushima nuclear disaster in 2011, the German federal election in 2009, or the FIFA World Cup 2006. In this talk, I present our recent work to create methods that facilitate extraction of event – centric sub-collections from the Web and Web archives. Creation of sub-collections raises several challenging research questions with respect to the crawler guidance, indexing and relevance estimation. On the Web, our methods are facilitated through social media guidance using Twitter and enable efficient monitoring, gathering and analysis of the fresh online content regarding current events. In Web archives, we propose flexible re-crawling methods coupled with topical and temporal relevance estimation and light-weight indexing. We discuss the opportunities and challenges of these approaches and present a framework for creating sub-collections.
21 January 2016. University of Bonn, Germany (Invited talk).
Title: Interactive Retrieval Methods, Sub-Collection Extraction and Semantic Alignment for Large and Heterogeneous Multilingual Datasets.
Abstract: The data available on the Web, in large-scale Web archives, in digital libraries and in open datasets is constantly growing and changing its appearance. The structural and linguistic heterogeneity of this data, as well as its large scale, substantially restrict its accessibility to the end users. In order to supply users with relevant and fresh multilingual information on demand, effective and efficient methods are essential that are able to cope with unknown data structures, large scale data and data written in foreign languages. In this talk, we explore methods tackling these challenges with the focus on: 1) Interactive retrieval techniques for structured data that do not require a-priori schema knowledge; 2) On demand extraction of event-centric sub-collections from the Web and Web archives to include fresh and relevant information; as well as 3) Semantic alignment of multilingual data to overcome the language barrier by creation of interlingual links.
Title: ALEXANDRIA – Analysing and Exploring Web Archives.
Title: Analyzing Relative Incompleteness of Movie Descriptions in the Web of Data: A Case Study. Wancheng Yuan, Elena Demidova, Stefan Dietze, Xuan Zhou. International Semantic Web Conference (Posters & Demos) 2014: 197-200.