Use this URL to cite or link to this record in EThOS:
Title: Geographical places as a personalisation element : extracting profiles from human activities and services of visited places in mobility logs
Author: Al-Azzawi, Ahmed N. Makki
ISNI:       0000 0004 2752 0321
Awarding Body: Cardiff University
Current Institution: Cardiff University
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
Access from Institution:
Collecting personal mobility traces of individuals is currently applicable on a large scale due to the popularity of position-aware mobile phones. Statistical analysis of GPS data streams, collected with a mobile phone, can reveal several interesting measures such as the most frequently visited geographical places by some individual. Applying probabilistic models to such data sets can predict the next place to visit, and when. Several practical applications can utilise the results of such analysis. Current state of the art, however, is limited in terms of the qualitative analysis of personal mobility logs. Without explicit user-interactions, not much semantics can be inferred from a GPS log. This work proposes the utilisation of the common human activities and services provided at certain place types to extract semantically rich profiles from personal mobility logs. The resulting profiles include spatial, temporal and generic thematic description of a user. The work introduces several pre-processing methods for GPS data streams, collected with personal mobile devices, which improved the quality of the place extraction process from GPS logs. The thesis also introduces a method for extracting place semantics from multiple data sources. A textual corpus of functional descriptions of human activities and services associated with certain geographic place types is analysed to identify the frequent linguistic patterns used to describe such terms. The patterns found are then matched against multiple textual data sources of place semantics, to extract such terms, for a collection of place types. The results were evaluated in comparison to an equivalent expert ontology, as well as to semantics collected from the general public. Finally, the work proposes a model for the resulting profiles, the necessary algorithms to build and utilise such profiles, along with an encoding mark-up language. A simulated mobile application was developed to show the usability and for evaluation of the resulting profiles.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available