A Semantic Association Based Literature Information Browsing System

Traditional key-word based literature retrieval and browsing is difficult to reveal large amount of information in various angles. Meanwhile, the information of which they pay attention to is single and simple, and they often ignore the recommendation of other important information and relations, in addition to literature itself. So, the efficiency and accuracy of retrieval and browsing is low. As a matter of fact, it is necessary to study what literature information to present and how to present them in a manner that facilitates researcher's retrieval and browsing.To tackle with these problems, in Semantic based Literature Association Information Browsing System– SLAIB, we present a Scientific Literature Semantic Association Model, and based on this model, an Association Information Visualization Strategy, Association Information Double Buffer Mechanism and related algorithm is adopted. We implement real-time interactive literature association information browsing in SemreX, which is a Semantic based Literature Retrieval and Sharing Platform. Scientific Literature Semantic Association Model depicts literature entities, explicit and implicit associations between entities, properties of entities, and properties of associations. Association Information Visualization Strategy, which adopt two dimensional net visualization method, show author's associations interactively and display important association information of Literature Semantic Association Model in different visualization metaphor, and support interactive search and dynamic browsing of association information. In order to ensure the real-time interaction of visualization, Association Information Visualization Strategy amalgamation important association information of Literature Semantic Association Model, and buffer them in both server and client using chain structure and recursion method. In addition, Author Recommendation algorithm, which follow the idea of"the bigger of the average value of related degree of two literatures between two authors, the more related the two authors are", compute author related degree of the Literature Semantic Association Model, so as to show the related degree in the visualization view, and lead user to browse author information that may be useful to the user.This system is implemented using Java language. Function and performance test is done based on this system. The function test mainly focus on the function of the association information visualization, and compare the function with two famous literature retrieval system-CiteSeer and DBLP. The result show that the function is complete, the association information is rich and the interaction is strong. The performance test mainly focus on the Association Information Double Buffer Mechanism, and test its system performance and expansibility. Result indicate that Association Information Double Buffer Mechanism provide efficient support when single user request multi-association cell and when multi-user request simultaneity. When the number of multi-association cell that single user request is up to 50, the response time reduces 98.46%, when the number of multi-user requested simultaneity is up to 110, the response time reduces 99.71%.