Analysis and detection of SIM box
Over a past two decades, Telecom industry is growing and mode of communication is changing and advanced day by day for catering individual and corporate needs. With the growth and advancement of technologies, telecom frauds are the major concern for research area and delivering the cent percent revenue in the system. Telecommunications fraud is a problem that affects operators all around the globe and one of the most known frauds is illegal to bypass fraud which is used in International voice traffic, in order to avoid carrier charges and this causes opportunity loss of international interconnect usage charge (IUC) to operators and this is major concern for research scope and impacting the revenue at operator level. As a result, cellular operators around the globe lose billions annually. Moreover, SIM box compromises the cellular network infrastructure by overloading local base stations serving these devices. This paper analyses the fraudulent termination of international traffic so suggest statistical, conventional, modern approach for detection of sim box and processes hundreds of millions of anonymized voice call detail records (CDRs). Their outputs of these models are optimally fused to increase the detection rate of sim box. The operator’s fraud department confirmed that the algorithm succeeds in detecting new fraudulent SIM box.
Published by: Vipin Airn
Author: Vipin Airn
Paper ID: V4I3-1243
Paper Status: published
Published: May 8, 2018
Dissimilar metal welding of J4-16Cr. austenitic stainless steel with Grade 201LN austenitic stainless steel experimental analysis through optimization Taguchi method
A literature review of studies and research has been made in the field of dissimilar metal welding. It has various industrial applications such as in the field of automobile chemical mechanical thermal power generation, nuclear plant. The main purpose of this paper is to review (a) Dissimilar metal welds and aspect of mechanical and metallurgical point of view. (b) Enhancing GMAW Technique. And second, the main objective of present study is to apply the Taguchi method.In this technique, an orthogonal array, signal to noise ratio (S/N)and analysis to variance (ANOVA) are made to study the welding characteristics of material & optimize the welding parameters.
Published by: Harpreet Singh, Barkat Ali
Author: Harpreet Singh
Paper ID: V4I3-1278
Paper Status: published
Published: May 8, 2018
Automatic session generator
The manual system of preparing timetables in colleges with a large number of students is taking a lot of time and usually ends with different classes struggling with the same teacher in the same room or at the same time. it occurs. To overcome all these problems, an automated system is proposed to be created. The system will take various inputs like students, subjects and classroom rooms and details of teachers on the basis of these inputs, it will generate a potential timetable, so that optimum utilization of all resources will be done in such a way that any obstacle or rules of the college according to. The list of topics can include electives as well as main topics.
Published by: Shammi Nanda, Amit Kumar Sharma
Author: Shammi Nanda
Paper ID: V4I3-1295
Paper Status: published
Published: May 8, 2018
Comparative study on Apriori and FP growth algorithms in big data
A dataset is a collection of data. Today the huge amount of the data is being captured by information sensing devices such as mobiles, computers, sensors etc. These huge amounts of the data are now called as big data. Frequent Itemset mining is a tool for identifying the frequently occurring items together. There are many frequent Itemset mining algorithms like apriori, Eclat and FP growth. In the proposed work we use FP growth algorithm and compare it with Apriori algorithm and we show that FP growth algorithm is better when compared to the apriori algorithm.
Published by: Reena Lobo, Venkatesh
Author: Reena Lobo
Paper ID: V4I3-1318
Paper Status: published
Published: May 8, 2018
A worthwhile performance framework modeling hinge on lambda architecture for batch and stream big data
The amount of data we are now generating is astonishing. Data has also evolved dramatically in recent years, in type, volume, and velocity. The emerging technologies like smartphones and sensors present opportunities for data exploitation, streaming and collecting from heterogeneous device every second. Analyzing these large datasets can unlock multiple behaviors previously unknown, a help optimizes approaches to many applications. However, collecting and handling of these massive datasets present challenges in how to perform optimized the large data. There are several frameworks available for handling the big data applications. The Lambda Architecture is data processing framework that can handle both batch and stream processing. The batch layer is implemented using Pig and hive, the streaming layer is built by using the Spark streaming and Spark SQL. This presents a need for developing the new framework for handling the big data applications particularly using public clouds to minimize cost, resource availability.
Published by: Athira Soman, Smitha Jacob
Author: Athira Soman
Paper ID: V4I3-1281
Paper Status: published
Published: May 8, 2018
Protecting information by hiding sensitive data attributes
Data mining aims at extracting hidden information from data. The process of discovering useful patterns and relationships in the large volume of data is called data mining. The goal of the data mining process is to extract information from a data set and transform it into an understandable structure. It involves databases, data management aspects, visualization & online updating. Data mining poses a threat to information privacy. Privacy-preserving data mining hides the sensitive rules and prevents the data from being disclosed to the public. The objective is to propose a novel association rule hiding (ARH) algorithm to hide the sensitive attributes. A function is used to obtain a prior weight for each transaction, by which the order of transactions modified can be efficiently decided. Apriori is used to find the frequent itemset with minimum support and confidence. Sensitive rules are generated based on frequent itemsets and the FHSAR algorithm is used for hiding sensitive association rules. This paper analyses the dataset obtained from SPMF an open source data mining library which is prepared based on real-life data. This paper shows the effectiveness of the algorithm.
Published by: Sighila P, Sangeetha S
Author: Sighila P
Paper ID: V4I3-1286
Paper Status: published
Published: May 8, 2018
