E-Mail Spam Detection and Classfication Using SVM and Feature Extraction
Today emails have become to be a standout amongst the most well-known and efficient types of correspondence for Internet clients. Hence because of its fame, the email will be misused. One such misuse is the posting of unwelcome, undesirable messages known as spam or junk messages. Email spam has different consequences. It diminishes productivity, consumes additional space in mail boxes, additional time, expand programming damaging viruses, and materials that contains conceivably destructive data for Internet clients, destroy stability of mail servers, and subsequently clients invest lots of time for sorting approaching mail and erasing undesirable correspondence. So there is a need of spam detection so that its outcomes can be reduced. In this paper, propose a novel method for email spam detection using SVM and feature extraction which achieves accuracy of 98% with the test datasets.
Published by: Shradhanjali, Prof. Toran Verma
Author: Shradhanjali
Paper ID: V3I3-1608
Paper Status: published
Published: June 26, 2017
A Review E-Mail Spam Detection and SVM Classfication Techniques
Today emails have become to be a standout amongst the most well-known and efficient types of correspondence for Internet clients. Hence because of its fame, the email will be misused. One such misuse is the posting of unwelcome, undesirable messages known as spam or junk messages. Email spam has different consequences. It diminishes productivity, consumes additional space in mail boxes, additional time, expand programming damaging viruses, and materials that contains conceivably destructive data for Internet clients, destroy stability of mail servers, and subsequently clients invest lots of time for sorting approaching mail and erasing undesirable correspondence. So there is a need of spam detection so that its outcomes can be reduced. In this paper, we show different spam detection techniques
Published by: Shradhanjali, Prof. Toran Verma
Author: Shradhanjali
Paper ID: V3I3-1607
Paper Status: published
Published: June 26, 2017
Online Internet Knowledge Management System for College
This project is aimed at developing an online intranet knowledge management system that is of importance to either an organization or a college. The system (KMS) is an Intranet based application that can be accessed throughout the organization or a specified group or Department. This system can be used as a knowledge or information management system for the college. Students or Staff logging should be able to upload any kind of technical information. Students or Staffs logging in may also access or search any information put up by others. Knowledge Management System should facilitate knowledge sharing from the grass root level like project teams to the entire college or organization.The Knowledge Management System is a web based system,which provides a platform for students and the staff of an organization or institution to upload any kind of technical or course related data.Students or Staffs logging in may also access or search any information put up by others. Knowledge Management System facilitates knowledge sharing from the grass root level like project teams to the entire college or organization.
Published by: Vikas .R, Venkateshwar .A, Nikesh Reddy, Tilak Kumar, Ranjeet .H
Author: Vikas .R
Paper ID: V3I3-1603
Paper Status: published
Published: June 24, 2017
Faculty Information System(FIS)
This is real time java based web application for faculty management. Where it consists of semester wise department time table. After completion of every session they should update student attendance with respective subject. Where absent students will be informed to their parents via text message. The system will generate the attendance percentage automatically day by day. The faculty can shift their session to another faculty when they are not able to take classes or busy in some other activities and the requested notification can be sent to the faculty mobile number.
Published by: Barge Sandesh, P. Phaniram Prasad, Sree Ananth Kumar .P, Musam Ahmed, Soumya .N
Author: Barge Sandesh
Paper ID: V3I3-1597
Paper Status: published
Published: June 23, 2017
Contextual Operation Using Pair Wise Ranking and Cot for Recommender Systems
The interest for omnipresent data preparing over the Web has required the improvement of setting - mindful recommender frameworks equipped for managing the issues of data over-burden and data separating. Contemporary recommender frameworks outfit setting - mindfulness with the personalization to offer the most exact proposals about various items, administrations, and assets. Be that as it may, such frameworks run over the issues, for example, meagerly, chilly begin, and versatility that prompt to loose suggestions. The cutting edge setting displaying strategies as a rule regard settings as specific measurements like those of clients and things, and catch importance's amongst settings and clients/things. In any case, such sort of pertinence has much trouble in clarification, e.g., it is not instinctive that a client is more pertinent to weekday than end of the week. A few chips away at multi-space connection forecast can likewise be utilized for the setting mindful proposal, yet they have restrictions in producing suggestions under a lot of logical data. Roused by late works in normal dialect handling, we speak to every setting esteem with an inactive vector, and model the relevant data as a semantic operation on the client and thing. Furthermore, we utilize the logical working tensor to catch the basic semantic impacts of settings. For the relevant data of every client thing collaboration, the logical operation can be displayed by duplicating the working tensor with inactive vectors of settings. However a client thing collaboration results can be produced under particular logical data yet can't be yielded under other relevant circumstances. So we propose a pairwise positioning limitation on the logical data. Our pair-wise positioning limitation uncovers the relative data among various logical circumstances and can be utilized to further improve setting demonstrating. Besides, we propose the top-n suggestion. It is another huge estimation of recommender frameworks.
Published by: Sachin Madhukar Kolase, Prof. V. V Jagtap
Author: Sachin Madhukar Kolase
Paper ID: V3I3-1595
Paper Status: published
Published: June 23, 2017
Design and Implementation of HDLC Controller Using VHDL Code
High Level Data Link Control (HDLC) is a bit oriented full duplex data link layer transceiver. HDLC controller has a Flag register with 8-bit pattern of 01111110 which generates the state of HDLC protocol. This paper describes the HDLC controller design using VHDL code consists of 16-bit cyclic redundancy checker (CRC) and FIFO. FIFO used for transmits the data. HDLC controller implement on Spartan 6 Xilinx FPGA.
Published by: Payal Gupta, Pankaj Gulhane
Author: Payal Gupta
Paper ID: V3I3-1590
Paper Status: published
Published: June 23, 2017