Combining an Improved Authorities and Hubs Algorithm with a New Multi-layer Classifier in a Topic Search Engine
Hongbin Zhang, Guangli Li · International Journal of Advancements in Computing Technology · 2011
Computer education resources on web have several searching topics and several file forms, so topic search algorithm plays an important role in information retrieval process as well as a good classification algorithm. And then an improved topic search algorithm named Authorities and Hubs and a multi-layer classifier are designed to resolve topic searching problems and resources classification problems. With the aid of the Authorities and Hubs algorithm and the classifier, users can easily make the precise classification fusion by any topic or by any file forms. Finally, lots of experiments which have been done step by step show that the accuracy of the topic search algorithm reached to 61% and it remained stable while the accuracy of the classification algorithm reached to 78% as well as the recall accuracy of the search engine reached to 63%.