期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：1
页码：776
DOI：10.15680/IJIRCCE.2017.0501165
出版社：S&S Publications
摘要：This paper shows brief description about the Markovian Semantic Indexing approach. The existingsystem uses the Latent Semantic indexing or Probabilistic Latent Semantic Index model. With this LSI havingproblems like when using large scale collection of images causes in low performance and speed levels and the PLSIapproach also has problems like it is incomplete since provide no probabilistic model at the level of documents, leads toover fitting problems if there are too many parameters in the model and it’s not clear how to assign and how to assignprobability to a document outside of the training data. In the context of online image retrieval system a new approachhas been introduced called Markovian Semantic Index. This model is useful when annotation data per image havinglarge training set data and suitable particularly in the context of online image retrieval systems.
关键词：Latent Semantic indexing; Probabilistic Latent Semantic Index model; Markovian Semantic Index
