A Comparative Study of the Utility of Knowledge Bases for the Geospatial Enrichment of Social Media Data
Date
Authors
Hunke, Jacqueline
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In the last decade the use of social media has increased substantially. It has become a data-rich source from which to extract meaningful data, as each post that a user makes to social media contains information that can be extracted or derived. Inherent in some of this data is geographic information (e.g. coordinates, location names) and this information, once discovered, can provide useful spatial, temporal, and thematic context. However, deriving such contextual information often requires the use of external knowledge bases. Employing Semantic Web and Linked Data principles have enabled the large-scale mining of such information from social media. As recent studies have shown, using such knowledge bases can disambiguate social media data and enrich its derived analysis products. Motivated by these recent trends, this thesis conducts a meta-analysis of those recent studies and proposes a set of selection criteria that should be considered when selecting a knowledge base for extraction of geographic and other contextual information from social media data. The analysis also explores which knowledge bases, among the ones existing today, are the most commonly-used and evaluates them against the defined criteria set. Based on this, a set of recommendations are put forward for selecting the best suited knowledge bases for the geographical enrichment of social media data.
Description
Keywords
Knowledge base, Social media, Geospatial enrichment, Semantic web