Intelligent Information Systems
We Derive Value from Data and Information
- How to leverage information?
- How to find new topics and trends?
- How to derive insight from heterogeneous/unstructured data and information?
- How to allow a «natural» access to data?
- How can software link data automatically?
These are but a few of the questions that the Intelligent Information Systems (IIS) group of the InIT is working to answer. While the “data and information flood” is often discussed negatively, we see a great opportunity to leverage data and information using the right approaches – both at search-time, as well as during analysis.
The research group transfers insights derived from research and development into teaching for students of the computer science curricula. It offers modules such as “Information Engineering 1 (Information Retrieval)”, “Information Engineering 2 (Data Warehousing & Big Data)” and "Databases". The group is active in both national and international research projects of the EU framework programs.
Research Topics
The Intelligent Information Systems group develops solutions for a changing, data-driven world. It performs research at the intersection of databases (DB), information retrieval (IR), data engineering (DE), natural language processing (NLP) and machine learning (ML)
The group covers two main research lines:
Big Data and Nano Data
We solve challenging problems when working with a range of datasets from very small (nano data) to very large (big data), where the nature of the problems change drastically as we work on different scales:
Current research:
- Information retrieval for small document collections
- Machine learning for query optimization
- Artificial intelligence for data integration and cleaning
- Quantum databases and quantum machine learning
Data Understanding
As we strive for "intelligent" solutions to data-driven problems, classical information systems need to process data at a different level, interpreting it to gain important information. Both structured and unstructured data must be processed not on a mechanical, but on a semantic level - e.g. by using natural language processing and understanding. Data is ultimately connected through graph structures or made accessible via semantic search.
Current research:
- Natural language interfaces for databases
- Semantic search on entities
- Knowledge graph construction
- Question answering over knowledge graphs
- Stream analytics and event detection
- Information retrieval evaluation
-
Digital Health Zurich – A practice lab for patient-centred clinical innovation
Digital Health Zurich researches digital health solutions in the hospital context and implements them efficiently and with practical relevance. Core topics are Patient Reported Outcome Measure (PROMs), remote monitoring, integrated care and related technologies as well as empowerment of patients and staff. Our ...
-
INODE4StatBot.swiss – Application of new algorithms for automatic natural language translation into database query language SQL (NL-to-SQL)
The goal of this project is to apply the major algorithms developed in the EU-Project INODE (Intelligent Open Data Exploration, https://www.inode-project.eu/) to Swiss Open Data. The focus is on developing multi-lingual extensions for so-called Natural Language to SQL systems (NL-to-SQL) where a natural language ...
-
Smart UMH: Smart urban multihub concept: Sustainable and liveable cities with low logistics visibility
Cities suffer from too much traffic, leading to congestion, air and noise pollution. Increased e-commerce popularity intensifies these challenges further. The Covid crisis has proven that our urban logistics systems are neither reliable, resilient, nor sustainable. Our objective is to develop a future urban ...
-
Meier, Benjamin Bruno; Elezi, Ismail; Amirian, Mohammadreza; Dürr, Oliver; Stadelmann, Thilo,
2018.
Learning neural models for end-to-end clustering [paper].
In:
Artificial Neural Networks in Pattern Recognition.
8th IAPR TC3 Workshop on Artificial Neural Networks in Pattern Recognition (ANNPR), Siena, Italy, 19-21 September 2018.
Springer.
pp. 126-138.
Lecture Notes in Computer Science ; 11081.
Available from: https://doi.org/10.1007/978-3-319-99978-4_10
-
Hibraj, Feliks; Vascon, Sebastiano; Stadelmann, Thilo; Pelillo, Marcello,
2018.
Speaker clustering using dominant sets [paper].
In:
2018 24th International Conference on Pattern Recognition (ICPR).
24th International Conference on Pattern Recognition (ICPR 2018), Beijing, China, 20-28 August 2018.
IEEE.
pp. 3549-3554.
Available from: https://doi.org/10.1109/ICPR.2018.8546067
-
Amirian, Mohammadreza; Schwenker, Friedhelm; Stadelmann, Thilo,
2018.
Trace and detect adversarial attacks on CNNs using feature response maps [paper].
In:
Artificial Neural Networks in Pattern Recognition.
8th IAPR TC3 Workshop on Artificial Neural Networks in Pattern Recognition (ANNPR), Siena, Italy, 19-21 September 2018.
Springer.
pp. 346-358.
Lecture Notes in Computer Science ; 11081.
Available from: https://doi.org/10.1007/978-3-319-99978-4_27
-
Lukic, Yanick X.; Vogt, Carlo; Dürr, Oliver; Stadelmann, Thilo,
2017.
Learning embeddings for speaker clustering based on voice equality [paper].
In:
2017 IEEE 27th International Workshop on Machine Learning for Signal Processing (MLSP).
27th IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2017), Tokyo, 25-28 September 2017.
IEEE.
Available from: https://doi.org/10.1109/MLSP.2017.8168166
-
Stockinger, Kurt; Bödi, Richard; Heitz, Jonas; Weinmann, Thomas Oskar,
2017.
ZNS : efficient query processing with ZurichNoSQL.
Data & Knowledge Engineering.
2017(112), pp. 38-54.
Available from: https://doi.org/10.1016/j.datak.2017.09.004