Title: Dynamic data source selection for humanistic culture integrated information retrieval

Authors: Song Deng; Haojing Weng; Zhonghao Hu; Fan Zhou; Wenhua Wu

Addresses: School of Software and Communication Engineering, Jiangxi University of Finance and Economics, Nanchang 330013, China ' School of Software and Communication Engineering, Jiangxi University of Finance and Economics, Nanchang 330013, China ' School of Software and Communication Engineering, Jiangxi University of Finance and Economics, Nanchang 330013, China ' School of Software and Communication Engineering, Jiangxi University of Finance and Economics, Nanchang 330013, China ' School of Software and Communication Engineering, Jiangxi University of Finance and Economics, Nanchang 330013, China

Abstract: One of the core technologies of data source selection is to construct summary. When the data source content updates, it will inevitably lead to changes in the data source summary. The calculation work load to completely update the summary is huge, so a partial update strategy is used to improve the efficiency. In order to enhance the efficiency of tourism cultural information integration, we propose a new dynamic data source selection strategy. Firstly, we build a summary based on the relationship between people and characteristic words. Then, we design a local summary updating algorithm based on sampling technique. Finally, we propose a scoring method of data source base on the gain of cultural content of a single or synthetic character for integration purposes. We conduct a number of experiments based on field data collection and the results showed that compared with the existing update algorithms, our method has higher accuracy.

Keywords: data source selection; cultural integration; dynamic; information retrieval.

DOI: 10.1504/IJIMS.2018.095247

International Journal of Internet Manufacturing and Services, 2018 Vol.5 No.4, pp.365 - 375

Received: 26 Jul 2017
Accepted: 01 Dec 2017
Published online: 02 Oct 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article