PhD Studentship Place Name Disambiguation using Semantic Web Technologies

PhD Studentship Place Name Disambiguation using Semantic Web Technologies

Web & Internet Science

Location:  Highfield Campus
Closing Date:   Monday 06 June 2016
Reference:  733216FP

PhD Studentship Place Name Disambiguation using Semantic Web Technologies

Web Science and Geography

Supervisors: Professor Wendy Hall, Professor David Martin

Location: Highfield Campus

Closing Date: 06 June 2016

Reference:

PhD Studentship: Place Name Disambiguation using Semantic Web Technologies

Names, whether applied to people, places, organisations or other things are an important form of identity but one that is far from perfect: an entity can have multiple names, and the same name can refer to multiple entities. These issues do not simply exist with classes of entity but also across them: “Warwick” can refer to a person, a place and an organisation. Even within the same class ambiguity can exist.  Two or more places can have the same name: Hampshire UK and Hampshire Illinois; a place can have different names in the same language: Derry and Londonderry, or in different languages: München and Muechen (German), Munich (English), Monaco (Italian); and so on.   When vernacular place names are considered or the granularity drops to local names the complexity can increase significantly.   This inherent ambiguity creates problems in terms of determining the correct entity being referred to.

Being able to differentiate place names to correctly identify place is important in a whole range of contexts including emergency services, delivery services, national security and the military. Two important uses are the process of extracting place related information from text analysis, and responding to an emergency call from a member of the public. In both cases ambiguity can result in the wrong location being identified which can significantly increase the time required to locate the correct place.

In information systems terms place names are usually managed within digital gazetteers. Contemporary gazetteers may serve an adequate role in terms of acting as map indexes but they are mostly information poor, at best hierarchical in organisation and frequently simple linear lists; as a consequence they fail when matched to meet the needs of the growing demand for sophisticated location-based data analytics. 

It has been proposed that the use of ontologies and Semantic Web technologies will enable more capable disambiguation services through the provision of richer and more detailed descriptions and the ability to infer relationships as well as supporting explicit relationships been place and place name.   However, what research that has been carried out has been limited to the level of settlements and above and has not attempted to address the issues that emerge when local issues and vernacular names are encountered.

The PhD will investigate the manner in which a sophisticated and information rich gazetteer can be represented and semantically described using Semantic Web technologies, and how this structure can be used to aid disambiguation.  It will be necessary to understand the complexity of sub-urban geographies and the manner in which ambiguity can be introduced.  It will also be necessary to understand how ambiguity is currently resolved and how this may be enhanced and translated into services that enable automatic and semi-automatic disambiguation to take place.

International collaborator / industrial partners: This project is part funded by the Engineering and Physical Sciences Research Council (EPSRC) and the Defence Science and Technologies Laboratory (Dstl).  

Eligibility:  Only UK citizens are eligible.    

If you wish to discuss any details of the project informally, please contact: Professor David Martin, Email D.J.Martin@soton.ac.uk