Information Extraction using Incremental Approach
Journal Title: INTERNATIONAL JOURNAL OF COMPUTER TRENDS & TECHNOLOGY - Year 2014, Vol 10, Issue 3
Abstract
Data mining is playing vital role in text extraction as now a day’s large amount of data available in scientific research, biomedical literature and web data. Data retrieval using existing approaches use sequential approach to process the data. It suitable for one time processing whereas using this approach performance will prunes. whenever the new data is added to the existing information we need to reprocess the entire data to perform extraction and it consumes large amount of time as same the initial time of processing .If at all there is any frequent modification in the existing data, it will require large amount of time to reprocess .This scenario will be repeats same even new extraction of goal is required for the same existing data. There is a high demand in the information extraction but available method such as UIMA and GATE performs IE by file based approach will not use any relational database in the extraction process. Key challenge of data extraction for incremental data, we need to identify which part of the data is getting affected by the change of any component or goal .To achieve this large corpus data will be stored using special type of data storage and optimized queries for data retrieval. It requires more storage compare to existing approach but now a days storage size not a key requirement. New approach also introduces automated query generation based on available input data for efficient performance. This method will reduce ninety percent of processing time whenever there is any modification of data comparatively to existing approach.
Authors and Affiliations
T. Ramesh Chary , N. Naveen Kumar
Keywords
Related Articles
- EP ID EP94231
- DOI -
- Views 137
- Downloads 0
Improved Discretization Based Decision Tree for Continuous Attributes
The majority of the Machine Learning and Data Mining applications can easily be applicable only on discrete features. However, data in solid world are sometimes continuous by nature. Even for algorithms that will directl...
An Efficient Boundary Detection and Image Segmentation Method Based on Perceptual Organization
In this paper, we presents a novel method for detecting the boundaries of the object in outdoor images by using most common properties of the images such as perceptual organization laws. Here the proposed segmentation sc...
A Support Vector Machine and Information Gain based Classification Framework for Diabetic Retinopathy Images
Image mining is the process of applying data analysis and discovery algorithms over large volume of image data. It has especially become popular in the fields of forensic sciences, fraud analysis and health care, for it...
Adaptive Integration of P2P and Mobile-Ad-hoc-Networks by a Cross Modeled CHORD Protocol
We set DHT (Distributed Hash Table) centered P2P (Peer to Peer) program-Chord in to MANET (Mobile Ad-hoc Network) in this paper. Then, we propose a new routing modified scheme Cross Modeled Chord(CM-Chord) which bases on...
Depth Sensor Based Skeletal Tracking Evaluation for Fall Detection Systems
Falls are very common in elderly due to various physical constraints. Since falls may cause serious injury and even death, fall detection systems are very important, especially when the victim is alone at home or is unab...