Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.779300
Title: Urban place profiling using geo-referenced social media data
Author: Lai, Juntao
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Achieving a better understanding of urban places benefits a variety of applications, such as urban planning, marketing, tourism, etc., hence it is a long-standing research topic. With the development of internet and mobile technologies, increasing number of individuals are contributing their place-related local knowledge through geo-referenced social media (GSM). This provides us new opportunities to obtain an accurate, detailed and dynamic understanding of urban places in a convenient and efficient way. However, place is a vague and abstract concept that is perceived and communicated by people, therefore it is difficult to formalise place information in digital systems. Furthermore, there are many issues in the extraction, organisation and analysis of place-related information from the disorganised local knowledge in GSM data, such as extracting local place names, estimating vague boundaries, inferring and representing their characteristics, especially in the complex and dynamic urban context. This research first defines a concept of place profile, as a structural information set with four elements, name, location, activities and time. A framework integrating spatial, temporal and semantic analysis is then designed to extract and formalise the information in GSM data regarding to these place elements. Two different approaches of associating GSM data to places are developed and presented, corresponding to the spatial and platial perspectives: the location-led approach which associates data to pre-defined spatial units, and the innovative name-led approach that associates data to place names. It is hoped that the place-profiling theory, the research framework and technical innovations presented in this thesis, can help to advance the formalisation of place in modern digital information systems, and benefit future urban place related studies and applications.
Supervisor: Cheng, T. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.779300  DOI: Not available
Share: