Tag Recommendation Research And Impiementati On Based On Text Analysis

Knowledge management and Topic model are increasingly popular topic, here is a combination of both of them for a institute knowledge management project, and aim to research and implement a auto tag recommendation project, at last the calculation of topic model for some documents in knowledge management project, and these topic words will helpful for calculation of documents similarity, and it have several steps as follow:1. First researching Latent Semantic Analysis and Probabilistic Latent Semantic Analysis, which can give an idea of what is Topic modeling and also Latent Dirchlet Allocation beaded on them.2.Next do some deep research to LDA, from Mathematical foundations, such as Binomial distribution, Beta distribution, to Gibbs Sampling, and completed a simple LDA program by MatLab.3.Combine with KM finished the project of auto tag recommendation, and in the process, one modified the algorithm via segmentation of words for tag aim, other one reconstructed algorithm by keep one topic in one sentence and divided document into three "title","overview" and "body", at last got a satisfactory results.