Weka-STPM: a Software Architecture and Prototype for Semantic Trajectory Data Mining and Visualization
Enormous quantities of trajectory data are collected from many sources, such as GPS devices and mobile phones, as sequences of spatio-temporal points. These data can be used in many application domains such as traffic management, urban planning, tourism, bird migration, and so on. Raw trajectory data, as generated by mobile devices have very little or no semantics, and in most applications a higher level of abstraction is needed to exploit these data for decision making. Although several different methods have been proposed so far for trajectory querying and mining, there are no software tools to help the end user with semantic trajectory data analysis. In this article we present a software architecture for semantic trajectory data mining as well as the first software prototype to enrich trajectory data with both semantic information and data mining. As a prototype we extend the Weka data mining toolkit with the module Weka-STPM, which is interoperable with databases constructed under OGC specifications. We tested Weka-STPM with real geographic databases, and trajectory data stored under the Postgresql/PostGIS DBMS.
Document Type: Research Article
Affiliations: 1: Departamento de Informatica e Estatistica, Universidade Federal de Santa Catarina 2: Instituto de Informatica, Universidade Federal do Rio Grande do Sul 3: Departamento de Informatica e Estatistica, Universidade Federal de Santa Catarina and, Instituto de Informatica, Universidade Federal do, Rio Grande do Sul
Publication date: 01 April 2011