Review on different Text documents clustering techniques
Along with the rapid and fast development of the Internet, there is a prodigious increase in the use of data and information. The aggressive growth of data has led us to an information explosion era, where the data cannot be easily maintained. Also, there is an increase in the use of electronic data and the information is stored in electronic format in the form of text documents such as news articles, books, digital library and so on. Clustering of the text documents has become an important technology over the internet. Text Clustering is mainly described as the grouping of the similar documents a large collection of unstructured documents. Text document clustering is the most widely used method for generalizing a large amount of information. In this paper, we tried to compare some existing text document clustering techniques on the basis of few criteria like time, accuracy and performance.
Published by: Jyoti Verma, Neetu Verma
Author: Jyoti Verma
Paper ID: V4I3-1695
Paper Status: published
Published: June 2, 2018
A mapping study on test case selection based on nature-inspired algorithms
After delivery of software product for modification, in order to correct faults or for improving the performance or other attributes, we calculate software maintenance. For this there is the need for regression testing, regression testing is used to check that no upcoming errors have been found throughout the maintenance phase. The abundant number of test suites consist of some repetitions/redundancies as the same fault/error may be covered by many test cases. Hence, it is recommended/advisable to decrease/reduce the test suite. Test case selection is one of the techniques used for reducing the number of test cases by selecting only those test cases from test suite which can detect all those faults which were detected by the whole test. This paper calculates the execution/performance of two metaheuristic algorithm – Cuckoo search and Bat algorithm for selecting test cases. Performance evaluation deciding factors are no. of faults detected and execution time. Results are achieved by conducting experiments on a large scale.
Published by: Sumedha Raheja, Rajvir Singh
Author: Sumedha Raheja
Paper ID: V4I3-1599
Paper Status: published
Published: June 2, 2018
A novel system for analysis of soil properties
The paper describes a microcontroller based model to test the soil sample for five different parameters viz moisture, temperature, pH, resistivity, and conductivity built using ATMEGA32. The readings sensed are stored in an Excel sheet through a Visual Basic code. The saved readings will be read in MATLAB software and graphical representations will be done in MATLAB. Data is secured at the file level and also while accessing the folder through passwords so that there is no unwanted interference. Conditions are tested by using Fuzzy logic and desired conditions necessary for plant growth will be concluded. Also, the graphs can be viewed on an ANDROID operating mobile phone system wherein users can view the plots by using the application MATLAB mobile. GSM modem SIM900 is used to send the readings via SMS to the concerned person giving scope for communication. Through this project, we can study variations in soil properties, the concept of precision farming and sustainable agriculture, overcome the limitations in conventional techniques, analyze the plant growth conditions and include scope for communication.
Published by: Sneha Nargundkar, Prabhakar Manage
Author: Sneha Nargundkar
Paper ID: V4I3-1710
Paper Status: published
Published: June 1, 2018
Vibrational analysis of cracked cantilever beam
A damage is one of the vital characteristics of the structural analysis because of safety cause as well as the economic prosperity of the industries. Identification of faults in dynamic structures and components are a significant aspect of judgment creating about their overhaul and retirement. The existence of cracks which influence the performance of structures as well as the vibrational parameters like modal natural frequencies, mode shapes, modal damping, and stiffness. In the present work, the effect of crack parameters (relative crack location and crack depth) on the vibrational parameters of a cracked cantilever beam are examined by different techniques using numerical method, finite element analysis (FEA), using ANSYS. Finite Element Method has been accomplished to derive the vibration signatures of the cracked cantilever beam. The results obtained analytically are validated with the results obtained from the FEA. The simulations of FEA have done with the help of ANSYS software. Similarly, the Modal analysis is carried out using ANSYS software. Harmonic analysis is done to observe the resonating and anti-resonating peaks for the Un-cracked and cracked cases considered. The stresses induced due to crack are obtained from the static structure analysis done with the help of ANSYS software. The natural frequencies values obtained from all these approaches are observed and the final conclusions are drawn. With the increase in crack depth, there is an increase in natural frequency. When the distance of crack from fixed end is increased there is an increase in natural frequency.
Published by: Sarat Chandra Menda, Pushpa Ratnam Raju Jalli
Author: Sarat Chandra Menda
Paper ID: V4I3-1686
Paper Status: published
Published: June 1, 2018
Hybrid technique for better noise removal to enhance edge detection
Edge detection technique is used to detect the edge of the image. But before edge detection the image must be free from noise. The solution for removal of image and image recovery is image de-noising models which give better solution to remove the possible pixel noise from the target image matrix. The image data is taken in the form of matrix data (2-D, 3-D or N-D), which is processed in the different dimensions with various practices in order to remove (eventually fix) the noise pixels. In this paper, the weight of the pixel is taken and covariance is calculated in the 3x3 pixel blocks. This algorithm does the cross-validation and checks the best fitness value of the pixel. Then the image is enhanced by using enhancement technique. Then the edges of the image are taken by using edge detection technique. The performance of proposed model has been estimated under various experiments. The results are found improved for all of the dataset images on the basis of PSNR (peak signal to noise ratio) and SSIM (structural similarity) based parameters.
Published by: Gurinder Singh, Pankaj Sharma, Puneet Jain
Author: Gurinder Singh
Paper ID: V4I3-1673
Paper Status: published
Published: June 1, 2018
Review on image watermarking based on LBP transform technique
In this review paper, we describe a digital image watermarking technique applying local binary patterns (LBP). Local binary patterns are best known for their rugged texture defining capabilities and digital watermarking applied in proving the copyright of a multimedia content. In this work, we study an LBP synthesis or inverse LBP identical process and its suitability to the digital image watermarking. The LBP synthesis process varies the near pixels values so that the LBP measured from these pixels is the desired value that we want to synthesize. This procedure takes into account the requisition of a digital image watermarking such as robustness and imperceptibility to watermark removal attacks. Due to the character of LBP synthesis, it is necessary that only a few pixels of a provided block is changed to embed the watermark. The simulation output shows that the technique is rugged to the rotation, JPEG compression, and scaling attacks. This LBP synthesis procedure could also be utilized to watermark sensor data for justifying the ownership. We are positive that this work would progress to a new research direction invalidation of digital content. In this work, a binary watermark is embedded into the image blocks by changing the neighborhood pixels conferred to the LBP pattern. However, various image blocks can have the similar LBP pattern, which can lead to false detection in watermark extraction procedure. In different words, one can change the host image deliberately without affecting its watermark message. In addition, there is no encryption procedure before watermark embedding, which leads to another potential security issue. To explain its weakness, two special copy-paste attacks are suggested in this paper, and several experiments are organized to prove the effectiveness of these attacks. To solve these issues, an enhanced semi-fragile watermarking based on LBP operators is conferred.
Published by: Riddhima Prakash, Faseeh Ahmad
Author: Riddhima Prakash
Paper ID: V4I3-1712
Paper Status: published
Published: June 1, 2018
