Dissertation > Excellent graduate degree dissertation topics show

Research on Similarity Search for High-dimensional Data Objects in P2P Networks

Author: JiangLan
Tutor: ZhuMing
School: University of Science and Technology of China
Course: Network communication system and control
Keywords: High-dimensional data High - dimensional indexing Similarity search
CLC: TP391.3
Type: Master's thesis
Year: 2009
Downloads: 61
Quote: 0
Read: Download Dissertation

Abstract


In order to achieve efficient access to massive multimedia information in P2P networks , multimedia information processing and retrieval tool developed to become a hot topic of current research . Traditional relational database based on accurate data query has been unable to meet the application requirements of content-based multimedia information retrieval technology continue to make research progress . To effectively support the increasingly complex queries , content - based information retrieval for multimedia data such as text, images, audio , video , data indexing so leads to a very challenging research topics : How to achieve mass on P2P , the high - dimensional feature vector data ( extracted from the multimedia object , used Characterization of the contents of the multimedia objects ) similarity retrieval . Efficient similarity search for massive , high -dimensional data in P2P networks , carry out the following work : ( 1) the PLCID indexing methods ( the modified iDistance based on the Proximity Location Code ) , the method utilizes approximate vector method and the method has the advantage of a one-dimensional vector conversion , retrieval , using two layers of filtration process : First , one-dimensional numerical search by distance map ; Secondly , the approximate location coding . This method effectively reduced the need to search the range of a decrease of the distance between the number of calculations in the data object , and thereby improve the retrieval performance . (2 ) PLCID index - based approach , the high-dimensional data retrieval in structured P2P network . The experiments show that , compared the original iDistance index , about a 30% increase on time performance , reduced system overhead by about 20% . ( 3 ) on the network load balancing problem , proposed a balanced program based on sampling ; based load change and distribution costs , measure the load of nodes in the network . Experiments show that this method has good load balancing effect .

Related Dissertations

  1. Research and Implementation of Web Service Registry and Similarity Searching,TP393.09
  2. Research on Algorithm of High Dimensional Outlier Detection,TP311.13
  3. Research on Sequential Pattern Mining and Time Series Similarity Search,TP311.13
  4. Virtual screening and design for the SARS coronavirus main protease inhibitors,R914.2
  5. Study on Similarity Query over Time Series Data,TP301
  6. Medical Image Retrieval Technology Based on Multiple-inverted File,TP391.3
  7. The Application Research of Time Series Data Mining on Stock Forecasting and Analysis,TP311.13
  8. Research on Improved Subspace Clustering Algorithm,TP311.13
  9. The Description and Search of Chemical Structures in Scientific Documents,TP391.3
  10. Research on Projected Clustering Algorithm and Its Applications,TP311.13
  11. Smart Home Network Traffic Abnormally Detecting Research and Implementation,TP393.08
  12. Research of Modified Strategy for Dynamic Electrocardiogram Waveform Classification,TP391.41
  13. Semi-supervised Subspace Clustering Based on Space-level Constraint,TP18
  14. Research on the Content-Based Audio Retrival,TN912.3
  15. Research on Clustering Methods for High Dimensional Data and Their Applications,TP311.13
  16. Based on research and application of high-dimensional data clustering algorithm,TP18
  17. Research on Graph Similarity Search Based on Frequent Sub-patterns,TP391.3
  18. Research on Subspace Clustering Algorithms Based on Density,TP311.13
  19. Research and Implementation of P2P multimedia document search system,TP391.3
  20. Near Neighbor Index for SVM Retrieval,TP391.3

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Retrieval machine
© 2012 www.DissertationTopic.Net  Mobile