Trajectory retrieval with latent semantic analysis

The problem of trajectory similarity has been recently attracted research interest considerably, due to its importance in diverse fields. In this work, we study trajectory similarity by attacking the problem taking an information retrieval perspective. Trajectories are first decomposed by using a grid and each trajectory is mapped to a multidimensional space where Latent Semantic Analysis is applied. Distance measures like Euclidean distance or cosine distance are applied to process similarity queries (range queries, kNN queries). Performance evaluation results, based on reallife data sets, show the simplicity and effectiveness of the proposed scheme. Categories and Subject Descriptors H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval General Terms Algorithms, Performance Keywords Trajectories, LSA, query processing

• Type: Journal
• Year: 2008
• Where: SAC
• Authors: Apostolos N. Papadopoulos
• Keywords: Applied Computing, Information Retrieval Perspective, Latent Semantic Analysis, SAC 2008, Trajectory Similarity