A Method of Personalized Tag Prediction Based on Graph Structure
Huang Yuan · 2017
In the past few years, social tags and tagging systems have gained large momentum for service categorizing and indexing content on the Web. Tags are used freely, which leads a random correspondence between the tags and the services, which affects the performance of the tag in the search and applications. We propose a novel scheme for tag predicting based on graph, aiming to automatically sort the tags associated with a given API services according to their relevance to the service content using the theory of random walk. Tag sorting is applied to tag predicting of Web service. We evaluate our methodology through experiment using open data set. Our results show that our approach really boosts the performances of tag predicting.