Where academic tradition
meets the exciting future

Data Mining and Knowledge Management Laboratory

Research

The amount of data and text has increased considerably during the last ten years and we are already talking about a future data amount of peta (1015) or even yotta (1024) bytes on the Internet. Today, many organizations struggle with vast amounts of data. Worldwide, computers have turned into massive data tombs. It is possible to capture and store data, but it has become difficult to utilize it effectively and efficiently.

Our overall research goal is to search for, find, model and systemize/analyze knowledge in very large sets of data using data- and text mining, so that organizations can use this knowledge in decision making.

Systematizing knowledge using data and in particular text mining is rather new and demanding. Focus is on the following application areas

Other areas of interest are:

Education based on research

Regular advanced courses in IS:

Research Unit Web Page: https://research.it.abo.fi/research/data-mining-and-knowledge-management-laboratory

Leader of the unit

Barbro Back

Co-leader of the unit

Tomas Eklund

Senior Researchers

Tomas Eklund Dorina Marghescu Peter Sarlin Zhiyuan Yao Hongyan Liu Henri Korvela

Doctoral Students

Piia Hirkman Annika H. Holmbom Samuel Rönnqvist

Projects 

Market Basket Analysis and CRM 2.0

In this project, new models on customer behavior and customer profiling in the retail areas have been tried out. In particular, a new method the Self-Organizing Time Map (SOTM) has been used in modeling customer behavior on the department store data. The research has been expanded to investigate a niche in the data i.e. green consumers. Combined segmentation and response modeling has also been researched. Moreover, evaluated our the visual segmentation models developed have been evaluated using the department store data. The evaluation conducted showed that the models are feasible ant they have been taken partly into use in one of the case companies.

Funding: Liikesivistysrahasto 2013–2014.

Text Analytics for Financial Risk

Text data constitutes a source that is particularly rich but challenging to harness. In the study of financial risk and stability especially, there is a pressing need to explore new types of data to model and understand the financial system and its risks. To further this aim theory and tools from natural language processing, machine learning, network analysis and information visualization are combined. The text analytics focus bridges text mining as well as user and domain-related aspects, where a visual analytics approach that emphasizes human-computer cooperation through visual interactive interfaces plays a key role as well.

Publications 

Click here to see the full list of publications from the TUCS Publication Database

The latest updated publications:

Niko Schenk, Christian Chiarcos, Kathrin Donandt, Samuel Rönnqvist, Evgeny A. Stepanov, Giuseppe Riccardi, Do We Really Need All Those Rich Linguistic Features? A Neural Network-Based Approach to Implicit Sense Labeling. In: Nianwen Xue, Hwee Tou Ng, Sameer Pradhan, Attapol Rutherford, Bonnie Webber, Chuan Wang, Hongmin Wang (Eds.), Proceedings of the CoNLL-16 shared task, 41–49, Association for Computational Linguistics, 2016.