2015 – 2019


Automatische Aggregation, Kontextualisierung, Zusammenfassung und Reformulierung von Nachrichten und Meldungen

Das Projekt beschäftigte sich mit der Zusammenfassung von News-Artikeln für News-Aggregatoren. Dabei wurden Lösungen für extraktive und abstraktive Summarization basierend auf Heuristiken und unter Anwendung maschinellen Lernens auf ihre Eignung für den konkreten Anwendungsfall untersucht. Entsprechende Funktionalitäten wurden in einem Prototyp implementiert. Im letzten Drittel des Projektes wurden in einer Pilotphase Ergebnisse aus dem Bereich News-Summarization auf die Rechtsdomäne übertragen und adaptiert.

2015 – 2018


Character Mining and Generation

The hero, the villain, the servant, the mentor, and many more ... movie and drama continue to rely on a repertoire of archetypical characters. But what makes a character? The proposed project CHARMinG will develop and apply AI methods from text and sentiment mining, natural language processing and machine learning to identify, from electronic sources of fictional dialogues (movie scripts, transcripts, drama texts), a set of indicators that convey the core of the relational/functional features and personality of characters, thereby leading to the generation of more colourful and engaging virtual characters.

2017 – 2018

De-escalation Bot

AI for a better discourse in online news fora

OFAI together with the Austrian (online) newspaper Der Standard developed the De-Escalation Bot, a series of machine learning based classifiers which scan forum postings for ones which are expected to provide valuable contributions to a forum and therefore should be visible not only for a rather short period of time but should be accessible to all users in a forum. The classifiers help forum moderators to identify such postings from a plethora of posts being produced by the forum users. Pinning respective posts on top of the forum has shown to improve the discussion quality, reports Der Standard (article in German).

2017 – 2018


Acoustic Monitoring of Biodiversity

The aMOBY project will use results from mathematical harmonic analysis combined with machine learning to acoustically monitor biodiversity. Biodiversity refers to the variety and variability of life on Earth, a diversity that is severely endangered due to human-made threats like habitat destruction, introduction of invasive species, over-population and over-harvesting, and of course climate change resulting from pollution of the atmosphere. Biodiversity monitoring is the repeated observation or measurement of biological diversity to diagnose and quantify its status and changes. A great challenge for monitoring of biodiversity lies in the sheer amount of data which clearly requires a high degree of automation to work on a grand scale.

2014 – 2018

On High Dimensional Data Analysis in Music Information Retrieval

Learning in high dimensional spaces poses a number of challenges which are referred to as the curse of dimensionality. Music Information Retrieval (MIR), as the interdisciplinary science of retrieving information from music, is very often relying on high dimensional feature representations and models. Hub songs are, according to the music similarity function, similar to very many other songs and as a consequence appear in very many recommendation lists preventing other songs from being recommended at all. It is due to the property of distance concentration which causes all points in a high dimensional data space to be at almost the same distance to each other. This proposed project will explore existing and develop new approaches to deal with these problems by studying their effects on a wide range of methods in MIR, but also multimedia and machine learning. In particular we are planning to (i) study and unify rescaling methods to avoid distance concentration, (ii) explore the role of hubness in unsupervised (clustering, visualization) and supervised learning (classification) in high dimensional spaces.

2015 – 2018


Artificial Language Understanding in Robots

ATLANTIS attempts to understand and model the very first stages in grounded language learning, as we see in children until the age of three: how pointing or other symbolic gestures emerge from the ontogenetic ritualization of instrumental actions, how words are learned very fast in contextualized language games, and how the first grammatical constructions emerge from concrete sentences. This requires a global, computational theory of symbolic development that informs us about what forces motivate language development, what strategies are exploited in learner and caregiver interactions to come up with more complex compositional meanings, how new grammatical structures and novel interaction patterns and formed, and how the multitude of developmental pathways observed in humans lead to a full system of multi-modal communication skills. This ambitious aim is feasible because there have been very significant advances in humanoid robotics and in the development of sensory-motor competence recently, and the time is ripe to push all this to a higher level of symbolic intelligence, going beyond simple sensory-motor loops or pattern-based intelligence towards grounded semantics, and incremental, long-term, autonomous language learning.

2013 – 2017


Spatial Memory and Navigation Ability in a Physically Embodied Cognitive Architecture

The aim of this project is the development of a computational cognitive model of spatial memory and navigation based on the LIDA (Learning Intelligent Distribution Agent) cognitive architecture, integrated with the other high-level cognitive processes accounted for by LIDA, and physically embodied on a humanoid PR2 robot with the aid of the CRAM (Cognitive Robot Abstract Machine) control system. The LIDA cognitive architecture will be extended by a conceptual and computational, hierarchical spatial memory model, inspired by the neural basis of spatial cognition in brains.

2015 – 2016


Kommunikationsplattform zur Vernetzung von Wissen und Workflows in Unternehmen

In the project, an approach was developed and implemented to classify chat messages into dialogue acts, focusing on questions and directives (“to-dos”). Our multi-lingual system uses word lexica, a specialized tokenizer and rule-based shallow syntactic analysis to compute relevant features, and then trains statistical models (support vector machines, random forests, etc.) for dialogue act prediction. The classification scores we achieve are very satisfactory on question detection and promising on to-do detection, on English and German data collections.

2013 – 2016

Automatic Segmentation, Labelling, and Characterisation of Audio Streams

The goal of this project is to develop technologies for the automatic segmentation and interpretation of audio files and audio streams deriving from different media worlds: music repositories, (Web and terrestrial) radio streams, TV broadcasts, etc. A specific focus is on streams in which music plays an important role.

2012 – 2016


Synchronization and Communication in Music Ensembles

Interpersonal communication and the coordination and synchronization of actions are fundamental human capacities. People use these functions routinely in activities such as shaking hands, driving a car, playing sports, or playing music as part of an ensemble. To coordinate your actions with someone else’s, you must be able to predict how the other person is going to behave. Music ensemble performance provides a particularly interesting context for studying prediction and coordination because the synchronization between actions must be so precise. Since music is dynamic, or time-varying, ensemble musicians must make predictions about their co-performers’ behaviour as they play, relying primarily on nonverbal cues provided by their co-performers’ body movements, breathing, and sound. This research project investigates the mechanisms underlying musical synchronization in small ensembles, using a combination of perceptual/performance experiments and computational modelling techniques.