A fast and efficient semantic short text similarity metric
- David Croft, S. Coupland, J. Shell, Stephen Brown
- Computer ScienceUK Workshop on Computational Intelligence
- 31 October 2013
A novel Short Text Semantic Similarity (STSS) method is presented to address the issues that arise with sparse text representation, and is shown to be comparable to current semantic similarity approaches, LSA and STASIS, whilst having a lower computational footprint.
Where Are the Pictures? Linking Photographic Records across Collections Using Fuzzy Logic.
- Stephen Brown, David Croft, S. Coupland, J. Shell, Alexander von Lunen
- Computer Science
- 9 September 2013
Fuzzy-matching algorithms combined with semantic similarity techniques offer a way of finding potential matches between such items when standard ontology and corpus-based approaches are inadequate, in this case helping researchers for the first time match photographs held in different archives to historical exhibition catalogue records.
A hybrid approach to co-reference identification within museum collections
- David Croft, S. Coupland, Stephen Brown
- Computer ScienceIEEE Symposium on Computational Intelligence for…
- 16 April 2013
A fuzzy rule based approach for identifying similarities between records and to identify co-referent records across multiple heritage collections is discussed and the application of this approach to real world collections and records which demonstrates some promising early results.
Improving record matching across disparate historical resources.
An attempt to use approximate reasoning techniques to automatically identify co-reference between records in collections held across different institutions to improve Record Matching Across Disparate Historical Resources.
A fast geometric defuzzication operator for large scale information retrieval
- S. Coupland, David Croft, Stephen Brown
- Computer ScienceIEEE International Conference on Fuzzy Systems
- 6 July 2014
A novel implication and centroid defuzzification approach based on geometric fuzzy sets and systems is presented that requires fewer operations and results in a significant reduction in processing time in the authors' application.
Picture Perfect: Computational methods for matching historical photographic records across different collections
Semi-automated co-reference identification in digital humanities collections
- David Croft
- Computer Science
- 1 July 2014
This thesis presents a novel method for co-reference identification between digitised records in humanities collections that is capable of searching collections stored using differing metadata schemas and requires comparatively little manual intervention.
The application of colour FIRE to robot vision
The main application of the colour FIRE presented here is that of robot vision, in particular the application to robot design for the puck collect competition held annually in Vienna, Austria.
An effective named entity similarity metric for comparing data from multiple sources with varying syntax
- David Croft, Stephen Brown, S. Coupland
- Computer ScienceDigital Scholarship in the Humanities
- 26 August 2016
The presented metric is effective at comparing named entity data in and across syntaxless data schemas such as are often encountered in Gallery, Library, Archive, and Museum collections.
Methods for Mining Messy Real World Data: Co-reference Identification Using Fuzzy Logic
This paper describes an alternative approach to handling messy real-world data, using a constrained hierarchical clustering approach on top of a fuzzy inferencing system that can accommodate messy real world data.