I did my PhD on work involving relation extraction from free text and automatic ontology
building. The aim was to extract subject - predicate - object
triples and use them to construct a graph database, implemented
in the Jena
RDF triple store. I explored RDF querying on a large scale, and
the graph database derived from free text was combined with one
generated from relational database tables and with published domain ontologies.
For the dataset I worked with, the theoretical maximum size of the
triple store was in the order of 235 million triples, but through
various optimisation techniques it was reduced to a more
manageable 25 million or so, excluding schema relations containing the
graph data structure (i.e. the ontology).
The data came mainly from RCAHMS
(The Royal Commission on the Ancient and Historical Monuments of
Scotland), with some material from two of the other National
Collections of Scotland: NMS
(The National Museum of Scotland) and
(The National Library of Scotland).
My MSc dissertation was on the use of NLP as an aid to Visual
Information Retrieval. The goal was to explore whether existing image
retrieval systems could be improved by carrying out shallow analysis of
the captions associated with digital images.
My PhD supervisors were Prof Ewan Klein
and Dr Claire Grover.
I'm currently working on European and JISC funded projects and looking for funding to pursue my own research plans.
- MSc Dissertation: Image Retrieval Using Natural Language and Content-Based Techniques (pdf)
- M-PIRO project with RCAHMS: RCAHMS Demonstrator Report (pdf)
- PhD Proposal: Relation Extraction for Ontology Construction (pdf)
- Proposed Annotation for Entities and Relations in RCAHMS Data (July 2006 version)((pdf)
- Proposed Annotation for Entities and Relations in RCAHMS Data (revised and updated, December 2006)(pdf)
- Draft Dissertation document (DDD)(pdf)
- Relational Database to RDF Translation in the Cultural Heritage Domain , May 2008 (pdf) (Chap. 5 of PhD thesis updates)
- PhD Thesis: Populating the Semantic Web - Combining Text and Relational Databases as RDF Graphs, (submitted Nov 2008, examined March 2009) (pdf) (copy in Edinburgh Research Archive).
- A printed and bound version of my PhD thesis is available for sale, ISBN 978-3843381161 (eg at MoreBooks or at Amazon). Feel free to review it!
- Populating the Semantic Web with Relations from Text, poster prepared for NaCTeM workshop, 28-29 Oct 2009 (pdf)
- Turning legacy data into Linked Data, poster prepared for DEMOfest event, 24 Nov 2009 (pdf)
- PictureBox, EDTC Feasibility Study Grant Report, October 2011 (pdf)
- K Byrne and E Klein. Image Retrieval Using Natural Language and Content-Based Techniques. DIR 2003, 4th DutchBelgian Information Retrieval Workshop, Amsterdam. (pdf)
- K Byrne. Tethering cultural data with RDF. JUC2006, Jena Users Conference 2006, Bristol. (pdf)
- K Byrne. Nested Named Entity Recognition in Historical Archive Text. ICSC2007, IEEE International Conference on Semantic Computing, Irvine, California. (pdf)
- K Byrne. Having Triplets - Holding Cultural Data as RDF. IACH2008, ECDL 2008 Workshop on Information Access to Cultural Heritage, Aarhus, Denmark. (pdf)
- K Byrne and E Klein. Automatic Extraction of Archaeological Events
from Text. Computer Applications in Archaeology, CAA 2009 (printed proceedings), Williamsburg, Virginia. (pdf)
- K Byrne. Putting Hybrid Cultural Data on the Semantic Web. Journal of Digital Information (JoDI),
10(6). Special issue on Information Access to Cultural Heritage. Eds.
Martha Larson, Kate Fernie, John Oomen. ISSN: 1368-7506. (pdf) (html)
Richard Tobin, Claire Grover, Kate Byrne, James Reid and Jo Walsh. Evaluation of georeferencing. In Proceedings of the 6th Workshop on Geographic Information
Retrieval (GIR'10), Zurich, Switzerland, Feb 2010. (pdf)
- Claire Grover, Richard Tobin, Kate Byrne, Matthew Woollard, James Reid, Stuart
Dunn, and Julian Ball.
Use of the Edinburgh Geoparser for georeferencing digitised
Philosophical Transactions of the Royal Society A, vol 368, no 1925, pp 3875-3889, August 2010.
Leif Isaksen, Elton Barker, Eric C. Kansa and Kate Byrne. GAP: A NeoGeo Approach to Classical Resources. Leonardo Transactions, May 2011. (pdf)
Leif Isaksen, Elton Barker, Eric C. Kansa and Kate Byrne. Googling Ancient Places. In Proceedings of Digital Humanities 2011 (DH2011), Stanford, CA, June 2011. (online paper)
- MSc presentation, Sept 2003
- DIR 2003 workshop, Dec 2003
- Presentation to DB Scotland group, Feb 2005
- Firbush study trip, Oct 2005
- PhD thesis proposal, Jan 2006
- JUC2006 conference, May 2006
- July 2006 progress report, July 2006
- Draft Dissertation Defence, Sept 2007
- ICSC Conference, Sept 2007
- Guest lecture for MASWS course, on "Relational Databases and RDF", Feb 2008; updated version, Jan 2009
- IACH Workshop, Sept 2008
- Glasgow Interoperability Conference talk, Scottish Government, Nov 2008
- SSP talk, "Populating the Semantic Web with Historical Text", Dec 2008
- Pre-viva presentation, March 2009
- CAA2009 conference, March 2009
- Guest lecture for MASWS course, on "Natural Language and the Semantic Web", Feb 2010; updated version, Feb 2011
- GIR'10 presentation, Feb 2010 (pdf)
- Presentation, "Text to RDF - and thence to the Web of Data" for Open Knowledge Scotland event, May 2010 (Keynote version or pdf)
DH2011 presentation (by Leif and Elton), June 2011 (pdf)
Guest lecture for MASWS course, on "Constructing RDF Datasets", Feb 2012; updated version, Feb 2013
- NeDiMAH workshop at
DH2012 July 2012, presentation and abstract