Probabilistic Recurrent Neural Network for Topic Modeling
P. Lakshmi Prasanna1, D. Rajeswara Rao2
1P. Lakshmi Prasanna, Koneru Lakshmaiah Education Foundation Vaddeswaram, Guntur (Andhra Pradesh), India.
2Dr. D. Rajeswara Rao, Koneru Lakshmaiah Education Foundation Vaddeswaram, Guntur (Andhra Pradesh), India.
Manuscript received on 05 February 2019 | Revised Manuscript received on 13 February 2019 | Manuscript published on 28 February 2019 | PP: 165-168 | Volume-8 Issue-4, February 2019 | Retrieval Number: D2673028419/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Data storing, and retrieving is the most important task in the current situation. Storing can be done based on the topic that the document describes. To know the topics, we have to classify the documents, to classify we are using topic modeling. In this paper we proposed probabilistic recurrent neural network (PRORNN) gives the most prominent result in the classification. it’s a Recurrent neural network (RNN)-based language model designed to directly capture the worldwide linguistics which means relating words during a document via latent topics. owing to their consecutive nature, RNNs square measure smart at capturing the native structure of a word sequence – each linguistics and syntactical – however would possibly face problem basic cognitive process long-range dependencies. As recurrent neural network fails to remember large dependencies, we are using topic modeling merged with probabilistic recurrent neural network which is called PRORNN. This PRORNN consists of all the merits of RNN and latent topic models. Thus, it gives most accurate classification as the result. The proposed PRORNN model integrates the merits of RNNs and latent topic models. In this paper we take the 20 news groups data set in that we take 2000 documents and we can labeled to two topics. to classify this 2000 documents and assigned 2 topics to for that documents and use the rnn package to execute recurrent neural network in R Tool.
Keyword: PRORNN, Classification, Topic Modeling, Local, RNN.
Scope of the Article: Neural Information Processing