Data Mining and Knowledge Discovery Research

Victoria Hodge is a Research Associate focusing on Data Mining and Knowledge Discovery. My main areas of research are Information Retrieval; Spell Checking; Classification, Categorization, Prediction and Estimation (including attribute selection and attribute weighting); Outlier and Anomaly Detection; Pattern Matching; and, Binary Neural Networks.

Victoria is a member of the Advanced Computer Architectures Group, Department of Computer Science at the University of York, UK.

We have developed an Information Retrieval system that incorporates the Information Retrieval process into a modular Neural Network architecture. The system uses binary, associative-memory neural networks to implement spell checking, synonym retrieval and word-document indexing.   Information Retrieval ...

We have developed a classifier/predictor using a binary, associative-memory neural network which implements the k-nearest neighbour algorithm. The system is designed for scalability and fast performance. Our research in this area also encompasses attribute selection and attribute weighting for use with classification and prediction.   Data Mining and Classification ...

Projects

We are currently developing an Intelligent Transport System within the FREEFLOW project which is funded by EPSRC, DfT and TSB. The system will detect patterns and anomalies in traffic flows, use pattern matching to find similar historical patterns and then take suitable measures to improve the current traffic flow using both historical knowledge and traffic modelling.   More Info...

Publications

My publications are listed on my Publications page. They cover: Information Retrieval; Spell Checking; Classification, Categorization, Prediction and Estimation; Outlier and Anomaly Detection; and, Pattern Matching and Binary Neural Networks.  More Info...

Last updated: 25 Mar 2013