My scientific passion lies in inventing and developing tools that help users in navigating their way through vast amounts of data.
Eva earned her master’s degree in Computer Science at the University of Innsbruck and subsequently pursued her PhD from the University of Innsbruck in the field of recommender systems for collaborative social media platforms. Her main research interests are within the fields of social media analysis, recommender systems and information retrieval. Over the last years, she has combined these three fields of research and investigated music recommender systems based on data retrieved from social media platforms aiming to exploit new sources of information for recommender systems. She was awarded a Postdoctoral Fellowship for Overseas Researchers from the Japan Society for the Promotion of Science allowing her to make a short-term research stay at the Ritsumeikan University in Kyoto.
- Development and application of concepts, methods and visual tools for interactive presentation and analysis of archivals (funded as a Lighthouse-project in the Field of Digitization by the Tyrolean Regional Government; 2018-2020; 200k €, Co-Principal Investigator)
- GoStudent (funded by the Austria Research Promotion Agency (FFG); 2019-2020; 90k €, Principal Investigator)
- various smaller projects in cooperation with industry (funded by the Austria Research Promotion Agency (FFG); 2013-2019; total of 30k €, Principal Investigator);
- Tweet-Models für Empfehlungssysteme in Mikroblogs (funded by the University of Innsbruck; 2014, 17k €, Principal Investigator)
- Hashomender Hashtag Recommendations in Twitter (funded by the Tyrolean Funds for Science; 2012; 5k €, Principal Investigator)
- Recommender Systems for the Creation and Maintenance of Structure within Modern Information Systems and Networks (funded by the University of Innsbruck; 2011-12, 8k €, Principal Investigator)
In our research, we are particularly interested in the following research areas:
Context-aware Music Recommendations
At the moment we are facing a fundamental change in the way people consume music: More and more people switch from private, mostly limited music collections to public music streaming collections containing several millions of tracks generating tons of data. The usability of such streaming services heavily relies on good recommender systems assisting users in discovering music they like. This makes the field of music recommendation and music information retrieval in a highly interesting topic for academia as well as industry. The DBIS Team focuses on context-aware music recommendation, exploiting data sources as Twitter, last.fm. or Spotify. Our research is concerned two types of context: Firstly, we focus on the current activity of a user while listening to music. Secondly, we are concerned with the cultural embedding of a user. In this research project we analyze music listening behavior using machined learning techniques. The generated insights are integrated into music recommender systems, aiming at improving their prediction accuracy.
Understanding Wikipedia Usage and Quality
Wikipedia has long become a standard source of information on the web and as such is widely referenced on the web and in social media. Articles on Wikipedia are created and maintained by a committed community, where individual users update, extend and review articles in a collaborative effort. Naturally, the quality of articles is of great importance to the community. Our research aims at making quality on Wikipedia measurable and investigate novel approaches to increase quality on Wikipedia. One example for a method for improving quality is the Wikidata platform, a crowdsourced, structured knowledgebase aiming to provide integrated, free and language-agnostic facts which are-amongst others-used by Wikipedias. Users who actively enter, review and revise data on Wikidata are assisted by a property suggesting system which provides users with properties that might also be applicable to a given item. We are particularly interested in improving this recommendation mechanism and hence, assisting users to contribute to an even more integrated, consistent and extensive knowledge base serving a huge variety of applications.
Semistructured Data and Recommendations
Current mass-collaboration and social media platforms use tags (or key-value pairs) to annotate and categorize resources enabling effective search capabilities. Due to the broad span of users of such systems (originating from different cultures and backgrounds, speaking different languages, etc.), the information provided features a limited amount of common structure, as e.g., objects are named differently and information is structured differently. This is a severe constraint in regards to the performance of search facilities. Our research aims to advance current recommendation approaches for semi-structured data to overcome structure heterogeneity and to develop tools to support users in the process of information curation.
Microblog Analyses and Recommendations
Despite the huge volume of tweets posted, this data hardly features structure in terms of categorization of tweets. The only structural information available are so-called hashtags which are a means to add simple keywords as a part of the tweet. However, as hashtags may be chosen freely by the users, the hashtag vocabulary is heterogeneous. Searches for hashtags in order to find tweets concerning a certain topic may result in a search result featuring low recall due to this heterogeneity of hashtags. Our research focuses on the development of recommender systems to support users by providing recommendations for suitable hashtags. Such a recommender system aims to add structure to microblog entries and hence, a more homogeneous set of hashtags enabling better search performance.