Projects
MoToRes: MoToRes: Mobilität und Tourismus für individuelle Nutzerbedürfnisse und regionale Spezifika (BMDV: 2024-2027)
Anforderungen an Mobilität und Tourismus (M&T) werden von Nutzerbedürfnissen, Umweltbewusstsein und den Zielregionen bestimmt. MoToRes soll Reisenden basierend auf Künstlicher Intelligenz (KI) eine individuelle, kontextabhängige, umweltbewusste und verständliche Beratung ermöglichen, um so die Nachhaltigkeit, Auslastung der M&T-Infrastruktur und die Verschiebung im Modal-Split für M&T zu stärken.
RUSHMORE – Resources for Human Mobility Research (DFG: 2024-2027)
Understanding mobility behavior, for instance, traffic flows or public transportation usage in urban areas, has become an increasingly important challenge for a wide variety of stakeholders, including social and environmental scientists, urban planners, federal statistical agencies, and policymakers. These stakeholders are particularly interested in understanding the short-, mid-, and long-term evolution of mobility behavior in specific geographic regions and its interdependencies with other factors. Such research requires high-quality data regarding the actual mobility behavior, as for example, historical archives of vehicle traffic data, shared bike usage, and public transportation usage. RUSHMORE will address sparsity and costliness of mobility data through representative synthetic data generation and specialized machine learning models for spatio-temporal knowledge transfer.
iBehave: Algorithms of Adaptive Behavior and their Neuronal Implementation in Health and Disease (NRW State Government: 2022-2026)
The goal of iBehave is to develop artificial intelligence methods to facilitate a better understanding of how the brain controls behaviours and facilitate a more accurate diagnosis and prediction of neurological diseases. DSIS research group participates in the interdisciplinary, intersectional project area IP7, aimed to uncover key mechanisms and develop machine learning methods to predict occurrences of epileptic seizures.
ATTENTION!: Artificial Intelligence for Trade-based Money Laundering Detection (ITEA, BMWK: 2022-2025)
IIllegaler Handel ist ein Problem von massivem Ausmaß, doch die Aufdeckungsrate liegt weltweit nahe null. Derzeit gibt es nur sehr wenige Werkzeuge, um komplexe, groß angelegte Fälle von Geldwäsche im Handelsverkehr und den Handel mit gefälschten Produkten aufzudecken. Das Projekt ATTENTION! wird vielfältige globale Datenquellen systematisch erfassen, semantisch modellieren und umfassend analysieren. Basierend auf globalen Handelsaktivitäten und deren Kontextinformationen werden Modelle der Künstlichen Intelligenz entwickelt und eingesetzt, um illegale Handelsaktivitäten und deren Muster in globalen, heterogenen Daten zu erkennen und aufzudecken.
KOALA: KnOwledge grAphs for disruption risk identification in suppLy chAins (DAAD, 2022-2023)
Weaknesses of the supply chain and associated risks are a severe threat to organizations and may lead to market cap and loss of revenue. Therefore, timely and effective identification of such risks is of utmost practical importance. Existing risk identification methods mainly rely on information about suppliers' actual operational performance. These methods do not proactively infer new risks based on external events early enough to minimize the impact of such events on the key operational performance indicators. The failure to capture risks from external information sources results in organizations in charge not having a complete picture of the uncertainties and experiencing severe negative impacts. The main challenge to improve this situation is identifying disruption risk events that may negatively impact supply chains timely and effectively using external sources. This project aims to investigate how approaches based on reinforcement learning and semantic knowledge representation in knowledge graphs can break the new ground in this context to facilitate timely and effective identification of supply chain risks.
WorldKG: World-Scale Completion of Geographic Knowledge (DFG, SPP VGIscience, 2019-2023)
OpenStreetMap (OSM) is a rich source of openly available volunteered geographic information. However, representations of geographic entities in OSM are highly diverse and incomplete. Recently emerged knowledge graphs (i.e. graph-based knowledge repositories) such as Wikidata, EventKG, and DBpedia provide a rich source of contextual information about geographic entities and support semantic queries. Whereas knowledge graphs provide a wide range of complementary semantic information for geographic entities, highly useful for Web applications, identity links between OSM and knowledge graphs are still rare and are mainly manually defined by volunteers. The main goal of the WorldKG project is to facilitate world-scale interlinking of OSM datasets describing different geographic regions with knowledge graphs such as Wikidata, EventKG, and DBpedia as well as completion of spatial knowledge in the knowledge graphs using OSM data.
Related publications: see bibsonomy.
smashHit: Smart Dispatcher for Secure and Controlled Sharing of Distributed Personal and Industrial Data (EU, H2020, 01/2020-12/2022)
The objective of smashHit is to assure trusted and secure sharing of data streams from both personal and industrial platforms, needed to build sectorial and cross-sectorial services, by establishing a Framework for processing of data owner consent and legal rules and effective contracting, as well as joint security and privacy-preserving mechanisms. The vision of smashHit is to overcome obstacles in the rapidly growing Data Economy which is characterized by heterogeneous technical designs and proprietary implementations, locking business opportunities due to the inconsistent consent and legal rules among different data-sharing platforms actors and operators. The Framework will provide methods and tools, such as Smart Data Dispatcher, to assure common consent over data shared using semantic models of consent and legal rules. The new tools include traceability of use of data, data fingerprinting and automatic contracting among the data owners, data providers, service providers, and users.
Related publications: see bibsonomy.
d-E-mand: Vorhersage von Ladebedarf bei Elektromobilität als Business Enabler (BMWi, 01/2020-12/2022)
Eine zentrale Voraussetzung für die Elektro-Mobilitätswende ist der Aufbau einer flächendeckenden Infrastruktur und datenbasierten Diensten für alle Arten von Elektrofahrzeugen. Im d-E-mand Projekt werden Lösungen entwickelt, um den Bedarf an der mobilen und stationären Ladeinfrastruktur, insbesondere bei durch Großveranstaltungen und Massenbewegungen (z. B. Ferienbeginn) verursachten Belastungsspitzen, systematisch zu prognostizieren und zu adressieren
Related publications: see bibsonomy.
CampaNeo: Plattform für Echtzeit Fahrzeugdaten Kampagnen (BMWi, 07/2019-06/2022)
Projektziel ist der Aufbau einer Plattform, auf welcher private und öffentliche Institutionen kampagnenbasiert und in Echtzeit Fahrzeugdaten erheben und analysieren können, sowie die Umsetzung von intelligenten Use Cases auf Basis der Kampagnendaten. Im Fokus stehen insbesondere die Data Ownerships der Fahrzeughalter sowie die Nachverfolgbarkeit der Datenverarbeitung.
Related publications: see bibsonomy.
Cleopatra: Cross-lingual Event-centric Open Analytics Research Academy (H2020-MSCA-ITN-2018, 01/2019-12/2022)
With a rapidly increasing degree of integration among the European countries, a rising number of events and topics strongly impact the European community and the European digital economy across language and country borders. This development results in a vast amount of event-centric multilingual information available from different communities and heterogeneous sources. This information can differ across the sources regarding its localization, potentially reflecting specific community-relevant aspects, containing cultural references and opinions as well as possibly incomplete or biased data. The main research objective of the Cleopatra ITN is to enable effective and efficient analytics of event-centric multilingual information spread across heterogeneous sources to deliver analytics results in the way meaningful to the users, with a particular focus on the journalists, digital humanities researchers and memory institutions.
Related publications: see bibsonomy.
Simple-ML: Big Data Machine Learning Workflows leicht gemacht (BMBF, 08/2018-07/2021)
Die effiziente Anwendung aktueller Machine Learning (ML)-Verfahren erfordert ein sehr hohes Maß an Expertenwissen, was einer verbreiteten Nutzung von ML-Ansätzen, insbesondere durch kleine und mittlere Unternehmen, im Wege steht. Das Ziel des Simple-ML Projekts ist daher die Benutzbarkeit von ML Verfahren signifikant zu verbessern um diese für einen breiten Anwenderkreis leichter zugänglich zu machen. Als zentraler Beitrag des Projekts wird eine domänenspezifische Sprache (DSL) definiert, die ML Arbeitsabläufe (Workflows) und deren Komponenten ganzheitlich beschreibt und sich durch textuelle und graphische Editoren spezifizieren lässt. Weiterhin leistet das Projekt Beiträge zur Robustheit der erstellten ML Workflows, Erklärbarkeit und Transparenz der erlernten Modelle, Effizienz und Skalierbarkeit der erstellten Anwendungen, sowie zur Wiederverwendbarkeit der erstellten Lösungen. Dies geschieht durch Anwendung semantischen Technologien, Weiterentwicklung von symbolischen ML-Verfahren und Aufbau auf skalierbaren ML-Frameworks. Die Ergebnisse des Simple-ML Projekts werden in den Anwendungsszenarien „Mobilität in der Stadt“ und „Logistik“ gemeinsam mit Anwendern aus der Wirtschaft validiert.
Related publications: see bibsonomy.
Data4UrbanMobility: Data Analytics for Mobility Services in Smart Cities (BMBF, 03/2017 – 04/2020)
Data4UrbanMobility focuses on facilitating innovative mobility services and mobility-related infrastructure development in smart cities through comprehensive data analytics. The stakeholders of the Data4UrbanMobility project are city councils, mobility service providers, and city inhabitants. The methods and tools developed in the Data4UrbanMobility project aim to provide insights into the mobility demand in smart cities, facilitate efficient use of the existing mobility services and infrastructure, support the development of innovative mobility-related services as well as facilitate effective planning of the mobility-relevant city infrastructure. To achieve this goal, the Data4UrbanMobility platform will interlink and enrich heterogeneous data sources including regional data collections, open data, and social media data using targeted Information Extraction, data integration, and machine learning methods. The methods and tools developed in the Data4UrbanMobility project will be validated in pilot projects in Region Hannover and Wolfsburg.
Related publications: see bibsonomy.