Multi-layer Reading Expansion System In Digital Library

Digital libraries have made rapid development and received a high degree of attention in many countries, which have become one of the most important portals for people to access information. Reading expansion system is one of the basic services in digital library. This paper studies the reading expansion problem in digital library and provides an efficient way for people to utilize the massive amounts of resources in digital library and enrich their own knowledge.The existed reading expansion systems are mainly based on artificial classification system, which require a large number of qualified specialists to spend a lot of effort on organizing the resources. However, all of these systems face the same problems such as anachronistic, limited coverage, coarse-grained and can't adapt to the user's own quest for knowledge requirements.The main work of this paper is as follows: First, capture book-review from Internet, apply book-centric diverse random walk algorithm on these high-quality articles, and provide topic-related and diverse keywords. Second, analyze CADAL's reading logs, track user's reading history, determine user's interest and recommend interesting keywords to reader through history-based book page keywords recommending algorithm. Third, find the most relevant words of query based on CADAL's catalogs, Expand the query and improve the performance of Lucene-based catalog search engine, link keywords to target knowledge. Fourth, develop the multi-layered reading expansion system and deploy it on CADAL as service.