Abstract: In mobile networks, reading requests and interests have high birth-death frequency and complex data structures, which are difficultly monitored and utilized by the server of library. In order to deal with them, a reading resource recommendation model was proposed based on mobile scene fusion. And its frameworks, working flows and key sub-algorithms were given. It used the mobile scenes as reading atlas nodes, the relations of mobile scenes as weighted reading atlas edges, and the grey fuzzy to integrate the library collection and reading requires, and customized the reading resource recommendation catalog. The experimental results showed that the model had higher client request coverage, better recommendation matching degrees and cost performance.

Key words: library service