Efficient Deep Learning Approach for Dimensionality Reduction using Micro blogs from Big data

Abstract

Nowadays Information Technology plays a vital role in every aspects of the human life. Now a world, the huge amount of stored information has been enormously increasing day by day which is generally in the unstructured form and cannot be used for any processing to extract useful information. Exploring potentially useful information from huge amount of textual data produced by micro blogging services has attracted much attention in recent years. An important preprocessing step of micro blog text mining is to convert natural language texts into proper numerical representations. Due to the short-length characteristics of micro blog texts, using term frequency vectors to represent micro blog texts will cause “sparse data” problem. Finding proper representations of micro blog texts is a challenging issue. In this project, we apply deep learning networks to map the high-dimensional representations of micro blog texts to low-dimensional representations. To improve the result of dimensionality reduction, we take advantage of the semantic similarity derived from two types of micro blog specific information, namely the retweet relationship and hash tags. Two types of approaches, including modifying training data and modifying the training objective of deep networks, are proposed to make use of micro blog-specific information. To improve the efficiency we implement the system in Hadoop. In addition to that to make services effective. To achieve the scalability and efficiency with help of map reduce framework in a big data environment.

Authors and Affiliations

Mr. M. Vengateshwaran, Mrs. C. Ramyapriyadarsini, Ms. N. Valarmathi

Keywords

Related Articles

Detecting Power Grid Synchronisation Failure on Sensing Frequency or Voltage beyond Acceptable Range

Our project is designed to develop a system to detect the synchronization failure of any external supply source to the power grid on sensing the abnormalities in frequency and voltage. There are several power generation...

Fusion based text detection and localization for video image

Text detection is much easier in images but it becomes complex with dealing with video images. Text is important for video indexing summarization and information retrieval text detection involves localization, extractio...

Nonlinear Pushover Analysis for Performance Based Engineering Design – A Review

Engineering Structures are designed to withstand all the anticipated loads without failure. Even after utmost care in assessing the expected loads as per the Codal provisions and field experience, modeling random loads...

Monopoly

Market is the most important place where seller and buyers meet each other and exchange commodity. In absence of market the economic growth of society goes down. Therefore a well-planned marketing is very important for...

Net Zero Energy Building

In this Report NZEB with providing Solar panels and other technics used. And It compared with conventional building. Its structure will be Net Zero Carbon Emission and to do researches to discover the most efficient pos...

Download PDF file
  • EP ID EP23199
  • DOI http://doi.org/10.22214/ijraset.2017.3002
  • Views 274
  • Downloads 6

How To Cite

Mr. M. Vengateshwaran, Mrs. C. Ramyapriyadarsini, Ms. N. Valarmathi (2017). Efficient Deep Learning Approach for Dimensionality Reduction using Micro blogs from Big data. International Journal for Research in Applied Science and Engineering Technology (IJRASET), 5(3), -. https://europub.co.uk/articles/-A-23199