A Novel Multi-Layer Sparse Regularizer based GRU Model for Consumer Reviews Summarization
Sourav SinhaResearch Scholar, Department of Computer Science & Engineering, B.S. Abdur Rahman Crescent Institute of Science and Technology sourav_cse_phd_17@crescent.education0000-0002-3327-8215
Dr. Revathi Sathiya NarayananProfessor, Department of Computer Science & Engineering, B.S. Abdur Rahman Crescent Institute of Science and Technology srevathi@crescent.education0000-0001-9584-5089
Dr. Indrajit MukherjeeAssitant Professor, Department of Computer Science& Engineering, Birla Institute of Technology Mesra imukherjee@bitmesra.ac.in0000-0001-8839-9981
Keywords: Text Summarization, Consumer Reviews, Sparse Regularization
Abstract
Text Summarization is considered as potential study area in recent past for its well-established growing popularity in domain specific areas. The consumer dominated society approved the summarization process as a tool for communication in shortest time span with wide coverage of valued basic knowledge. The Auto Text Summarization is developed digitally with various sophisticated tools and techniques in compressed opinion building process delivering millions of consumer’s verdict at one scratch. The newest methodology of deep learning neural network had been adopted to narrate the precise fact with proposed novel MSRGRU (Multilayer Stacked Regularizer GRU) model for consumer review summarization technique. The adaptation of the model by proper utilization of sparse regularization process enhances quality of predicted summaries. Eventually the accuracy level of the model is remarkable as compared to the existing state-of-the -art techniques.