A Novel Retrieval Method for Multimodal Point of Interest Data

Ying Xia, Shiyan Luo, Xu Zhang, Hae Young Bae
2014 International Journal of Multimedia and Ubiquitous Engineering  
With the rapid development of mobile Internet, mobile geographic information service becomes more and more popular. Point of interest (POI) is the core content of mobile geographic information service, and it exhibits the properties of multi-source, heterogeneity and media convergence under the circumstance of multiple users and service providers. At the same time, users' retrieval demand is becoming complicated and personalized. Traditional spatial data retrieval method and its semantic
more » ... ion cannot meet the needs of multimodal point of interest (MPOI) data access. In this paper, a novel retrieval method for MPOI which includes offline and online phases is proposed. In the offline phase, a two-step POI fusion algorithm and a Geo-ontology model were adopted to preprocess the MPOI data. In the online phase, a retrieval process is presented to do the practical retrieval. This method not only satisfies the requirement of multimodal fusion and semantic sharing in the mobile Internet, but also supports cross-modality multimedia retrieval. Experimental evaluations validate the performance of the proposed method. 291 as the evaluation criteria. According to several trials, the threshold of the One-sided Nearest Neighbor phase is determined as 0.0, and the threshold of LD algorithm is 0.6. Experiment 1: The performance comparison of two-step POI fusion algorithm and COM-NWT under different coincidence degree [15] .
doi:10.14257/ijmue.2014.9.7.24 fatcat:eocdfx7trbdtfcb3qoowpfdlyu