← Back to Archives
National Conference on Advancements in Information Technology
NCAIT-17
📅 Date: 19th & 20th May 2017
🏫 Organized by: JSS Academy of Technical Education, Bengaluru
📚 Department: Department of Information Science & Engineering
The Growing Influence of Predictive Analytics in Various Industries
Prashasthavardhana Prahlad, Sachin K. Rai, Dr. D.V. Ashoka
Opinion Analysis, Summarization and Timeline Generation for Dynamic Tweet Streams
Prateeksha M, Reshma J T, Sandesh S
An Offline Model to Overcome Online Theft by Considering Offline Payments
Akhilesh S Basri, Govind Prasad A, Karthik S
Associating Online Networking on E-commerce: Cold-Start Item Suggestion Using Microblogging Data
Tej Prakash Choudhary, Tejinder Singh, Vipul Jain
Dynamic Quad tree Mapping for Geographical Search with Keywords
Sindhushree C, Supriya S M, Suraj Wali
Smart and Secure Exam Management System for Mobile Learning Environment
Prashant Jha, Santosh Kumar Pandey, Sayantika Deb
Suraksha Digital Transformation to Indian Police Services
Meghana S, Hemashri R M, Aishwarya Nagesh Kamath
Wireless Body Area Network for Heart Attack Detection and Automatic Alert System
Pallavi Ganapati Gaonkar, Rishav Verma, Shashank K R
Abstract
The Growing Influence of Predictive Analytics in Various Industries
Prashasthavardhana Prahlad, Sachin K. Rai, Dr. D.V. Ashoka
Abstract:
Predictive analytics is the process of finding structure in historical data to come up with predictions about future events by using a combination of techniques from data mining, machine learning, and statistics. In this process, one usually constructs models to identify relationships among factors that potentially have an impact on future events. Predictive analytics is widely used in marketing, financial services, retail, travel and many other areas. The use of predictive analytics is growing rapidly, which can be attributed to the availability of new data sets and sources, increased computing power, and the constant, unrelenting search for business supremacy. One important fact is that, even though the transactions are taking place in various market sectors, it can be inferred that all the companies has the same strategic goal: to utilize the power of analytics to gain deeper insight into customer demeanor and to use that knowledge to better serve their customers. This paper will give more insight into the rapid expansion of predictive analytics in various markets.
Keywords:
Predictive Analytics, Data mining, machine learning, predictive modeling, statistics, markets.
Abstract
Analysis of Encryption Algorithms and Deduplication Technique in Cloud
Punashri M S, Nirmala H
Abstract:
The most well known and vital popular service of Cloud is Data Storage. Information are regularly put away in an Encrypted shape to secure and protect information security of holders. Traditional deduplication schemes cannot work on encrypted data. Existing solutions of encrypted data deduplication suffer from security weakness. They cannot flexibly support data access control and revocation.New difficulties for cloud information deduplication is presented from Encrypted information. In Cloud this ends up noticeably significant for information preparing and capacity. Existing arrangements of information deduplication have less security in others words security shortcoming. Repudiation and information get to controls. In this an answer for deduplicate encoded information in cloud in light of intermediary re-encryption and possession challenges. With get to control it coordinates cloud information deduplication. The execution assessment depends on Computer reproductions and broad investigation. Results show the better performance of encryption algorithms, and data deduplication in Cloud storage.
Keywords:
Cloud Computing, Encryption Algorithms, Digital signature, Deduplication.
Abstract
Opinion Analysis, Summarization and Timeline Generation for Dynamic Tweet Streams
Prateeksha M, Reshma J T, Sandesh S
Abstract:
Short-instant messages, for example, tweets are being made and shared at an exceptional rate. Tweets, in their crude shape, while being instructive, can likewise be overpowering. For both end-clients and information experts, it is a bad dream to push through a large number of tweets which contain gigantic measure of commotion and repetition. In this paper, we propose a summarization technique to lighten the issue. As opposed to the customary report outline strategies which concentrate on static and little scale informational index, the proposed summarization technique is intended to manage dynamic, quick arriving, and vast scale tweet streams. Our proposed system comprises of three note-worthy segments. In the first place, we propose an online tweet stream bunching calculation to group tweets and keep up refined insights in an information structure called tweet group vector (TCV). Second, we build up a TCV-Rank synopsis method for producing on the web rundowns and verifiable outlines of discretionary time lengths. Third, we plan a viable point advancement identification technique, which screens synopsis based/volume-based varieties to deliver courses of events consequently from tweet streams. Our trials on huge scale genuine tweets show the ef?ciency and adequacy of our system.
Keywords:
Tweet Cluster Vector, Summarization, TCV-Rank, Point Advancement Identification Technique.
Abstract
Text Detection from Images using Sharpening and Blurring Technique
Ajay B N, N Swetha
Abstract:
Text plays a pivotal role in the images. Text in images accommodate some useful raw facts about the scene and surroundings. Therefore, it is important to detect and recognize them. Some of the limitations to detect texts are intensity, fonts and sometimes even because of complex background. In order to overcome these limitations, first sharpen and blur the image using Gaussian blur and low pass filters, then all the covariant regions are detected. Using all the covariant regions, convolute the image with the Gaussian kernel and continuously down sample the image by half. After convoluting, create a feature map and apply morphological gradient to the resultant image in order to cluster the text regions. After doing this, the non-text regions are eliminated and the text regions are detected. The proposed approach is experimented on ICDAR 2015 dataset and the results show that the proposed approach is robust and it outperforms for various kinds of images.
Keywords:
Gaussian kernel, text detection, morphological gradient.
Abstract
Authentication using Images as Password for Providing High Security in Networks
Kavya H C
Abstract:
Graphical Password Authentication has been accepted and there is a significant increase in usage of it. As using Text based password is common throughout all the domain areas, it creates the situation where incase if a user selects a password that he can easily remember then an attacker is provided with an oppurtunity to easily guess. If user selects a password that an attacker could not guess, then it becomes hard to remember for the user. So, Graphical password scheme is used as humans can easily remember images compared to alphabets or numbers. The focus here is to replace the static images used in graphical password system with the tokens. Usually the server would store the images and display it for the user while log in. But this would provide an option for an attacker to easily know the password by using some hacking techniques. Here, the user is responsible to get the image that which he wants to make it as a password. Moreover, the image that he gives it as a password is not at all stored in the server. User has to give in his own image which he has selected and then choose some five locations in that. It is converted to some value and then stored in the server. In this way an attacker is restricted a little as he would never know which image the user has uploaded during registration and he would not have it as it would be the personal image of the user. Even though if an attacker somehow got to know what is been stored in server by using some attacks, he can get only converted value not full data about it. Usability, reliability, and security against observation is provided. The study shows that there would be a threshold set, only when the password items match with the originals it is said to be equivalent. Resistance to observation attack is increased by using this approach.
Keywords:
Graphical password, Security, Shoulder surfing, Authentication.
Abstract
An Offline Model to Overcome Online Theft by Considering Offline Payments
Akhilesh S Basri, Govind Prasad A, Karthik S
Abstract:
Online threats has been increasing due to the use of credit card and pos services. The customer data is got by consumer at this point of POS. A malware can be added to the pos device or a fault chip can be inserted into the POS to capture the customer details. This is a threat this can be overcome by considering the payment as offline. Here we present a offline model called OFFLINE MODEL that will be offline and payments will be through offline generating digital coins. This increases security in the payments.
Keywords:
POS, OFFLINE MODEL, digital coins.
Abstract
Associating Online Networking on E-commerce: Cold-Start Item Suggestion Using Microblogging Data
Tej Prakash Choudhary, Tejinder Singh, Vipul Jain
Abstract:
Numerous web based business sites bolster the component of social login where clients can sign on the sites utilizing their social networking identities, for example, their Facebook or Twitter accounts. Clients can likewise post their recently bought items on microblogs with connections to the web based business item site pages. In this paper, we propose a novel answer for cross-webpage item suggestion, which means to prescribe items from web based business sites to clients at informal communication destinations in "cold-start" circumstances, an issue which has once in a while been investigated some time recently. We propose to utilize the connected clients crosswise over person to person communication destinations and online business sites (clients who have interpersonal interaction accounts and have made buys on internet business sites) as a scaffold to guide clients long range interpersonal communication elements to another element portrayal for item suggestion. In speci?c, we propose learning both clients and items component from information gathered from online business sites utilizing neural systems and afterward apply a modi?ed inclination boosting trees strategy to change clients long range interpersonal communication highlights into client embedding.
Keywords:
Cold-Start; E-commerce; Microblogs; Neural Systems.
Abstract
Carbon Price of Cloud Computing.
D.V. Ashoka, Parthjeet, and Piyush Kumar
Abstract:
The digital cloud is thought to be invisible. Instead of physical storage devices like books, CDs, DVDs, newspapers, and magazines, we have pure data, traveling throughout the interconnected web. Now we can access everything at a push of a button. But is the cloud actually invisible? No, there is a physical substance to it: the data centers. The data centers space all over the world is big enough to fit approx. 6000 football fields. These data centers obviously run on electricity. And if the source of electricity is fossils, it does have a carbon price. As cloud computing is essential in modern practices, it cannot be compromised. Hence, a detailed analytical study is essential to bring down the negative impact of cloud computing on the environment. This paper overviews the effects of cloud computing on nature and advanced technologies which can be used to solve its negative impact.
Keywords:
Cloud, The Web, Data Centers, Carbon Price.
Abstract
Dynamic Quad tree Mapping for Geographical Search with Keywords
Sindhushree C, Supriya S M, Suraj Wali
Abstract:
Searching exact location of the hotels, shopping centres and cinemas in the specified area is a key challenge in web crawler systems. Nowadays various new applications are introduced for querying the objects and predicting associated documents. For example, instead of getting all the list of hotels and restaurants, user would be interested in searching for nearest neighbour list. Presently best solutions to such things are Quad tree mapping as proposed in this paper. Proposed solution consists of admin component which performs addition of hotels, shopping centres and cinema details in each area with geographical location. The user component searches for hotel, shopping centres and cinema in interested area. If requested area by the user is proper then Google map view will be displayed with exact location and distance from the user place. The proposed technique constructs Quad tree Mapping based on the distance in the descending order. User also gets the longitude and latitude of the place chosen. Efficient access method called inverted index is developed for experiment purpose of web application.
Keywords:
keyword search, spatial data, location based, quadtree, information retrieval.
Abstract
ISRO Awards Secure Web Application
Neelavarsha DK, Nikhil R, Nishchita A
Abstract:
The necessity of cyber security is more today than ever. A proliferation of cyber-attacks is causing increasing attacks to companies, institutions and the government. The numbers of threats are increasing evidently and organizations need to counter these attacks by using appropriate security measures. Cybercrime has steadily risen to become a major threat. According to the Global Economic Survey conducted in 2016, only 37 percent of organizations have a dedicated cyber incident response plan and these need to be driven to a greater extent. Security is important for two main reasons. Firstly, due to the increase in cyber-attacks in recent years and secondly, due to the severity of these attacks which are able to compromise the security and integrity of applications. The necessity of countering cyber threats is more important in private organizations and government organizations due to high precision and confidentiality of data. As the organizations become increasingly interdependent, analysts must pay more attention to the security of their organizations. The best way to counter attacks is the use of certain security plans and procedures.
Keywords:
Cyber security, Severity of attacks, precision, confidentiality.
Abstract
Nanotechnology in Health Care
Ishwarya Narahari, Aravinda B S, Dr. Malini M Patil
Abstract:
Nanotechnology deals with the behavior and manipulation of individual atoms and molecules whose size is less than 100 nanometers of a human genome. The break through work was announced in 1981 for the invention of scanning tunneling microscope and the discovery of nanocrystalline. Today�s scenario in nanotechnology is astounding as it has been integrated in many forms such as Nano agriculture, energy storage, water treatment and many more. Nanotechnology in the field health care offers even a wider extent of applications such as nanotubes, stretchable electronics and bleeding plastic. Nano medicine is the application of Nano technology to human health care by which the medical diagnosis of a patient and the relevant treatment can be efficiently managed. Another important parallel study about nanotechnology is the regeneration of organs and tissues. A large range of deadly diseases such as Alzheimer�s, cardio vascular diseases, neurological diseases can be treated very effectively and efficiently. The present work emphasizes on the study of concept of nanotechnology and its wide variety of applications with a case study.
Keywords:
Nano technology, nano tubes, bleeding plastic, genome, Nano mediciene.
Abstract
Predoctor Support
Priyanka G
Abstract:
Clinical archives are without rich content information sources containing important pharmaceutical and side effect data, which have an awesome potential to enhance human services. In this paper, we assemble an incorporating framework for removing prescription names and manifestation names from clinical notes. At that point we apply non negative lattice factorization (NMF) and multi-see NMF to group clinical notes into significant bunches in view of test highlight grids. Our test comes about demonstrate that multi-see NMF is an ideal technique for clinical record grouping. Also, we find that utilizing extricated prescription/side effect names to bunch clinical records beats simply utilizing words.
Keywords:
Clinical notes (records/ documents), document grouping, side effects or symptoms, prescription or medications, Multi-view, nonnegative matrix factorization.
Abstract
Mohan Kumar H U, Muralidhar K, Karthik B M
Abstract:
Postural instability affects many people and can compromise even many simple activities of the daily routine. The term Rapid Entire Body Assessment (REBA) depicts the usage and availability of postural assessment of jobs in numerous professional streams. As per the survey in major cases REBA methodologies have been used to compare the results with other observational units for level of conformity between the two when determined. This module is specifically designed to be sensitive to the type of unpredictable working postures found in health care and other service industries.
Keywords:
Postural analysis, accelerometer, andriod device.
Abstract
Review on Heterogeneous Network Technologies in Smart Grid
Samrin Banu, Sowmya K.N
Abstract:
In technological world smart grid helps in two way communication along with the transmission of the electricity on the same transmission line which strives to improve performance, security, reliabilities and lifetime of the heterogeneous network when deployed in building or industries. In this paper we discuss about architecture of smart grid, Technologies and standard protocols adopted which support the interoperability in smart grid and influence of heterogeneous infrastructure on improvising the latency, number of hops and life time.
Keywords:
Smart Grid, Power Grid, Network Performance, Lifetime, Mesh Network, Latency.
Abstract
A Robust and Reversible Watermarking Technique for Relational Data
Harish R, Vijay Kumar S
Abstract:
Advancement in information technology is playing an increasing role in the use of information systems comprising Digital data. The Digital Image data�s used effectively in collaborative environments for information extraction; consequently, they are vulnerable to security threats concerning ownership rights and data tampering. Watermarking is advocated to enforce ownership rights over shared data and for providing a means for tackling digital image data tampering. When ownership rights are enforced using watermarking, the underlying digital data undergoes certain modifications; as a result of which, the digital image data quality gets compromised. Reversible watermarking is employed to ensure data quality along-with digital data recovery. However, such techniques are usually not robust against malicious attacks and do not provide any mechanism to selectively watermark a particular attribute by taking into account its role in knowledge discovery. Therefore, reversible watermarking is required that ensures watermark encoding and decoding by accounting for the role of all the features in knowledge discovery and original encrypted watermarked data recovery in the presence of active malicious attacks In this paper, a robust and semi-blind reversible watermarking (RRW) technique for digital data has been proposed that addresses the above objectives. Experimental studies prove the effectiveness of RRW against malicious attacks and show that the proposed technique outperforms existing ones.
Keywords:
Digital data, RRW, Digital Image, reversible watermarking.
Abstract
Smart and Secure Exam Management System for Mobile Learning Environment
Prashant Jha, Santosh Kumar Pandey, Sayantika Deb
Abstract:
M-Learning has enhanced the e-learning by making the learning process learner-centered. However, enforcing exam security in open environments where each student has his/her own mobile/tablet device connected to a Wi-Fi network through which it is further connected to the Internet can be one of the most challenging tasks. In such environments, students can easily exchange information over the network during exam time. This paper aims to identify various vulnerabilities that may violate exam security in m-learning environments and to design the appropriate security services and countermeasures that can be put in place to ensure exam security. It also aims to integrate the resulting secure exam system with an existing, open source and widely accepted Learning Management System (LMS) and its service extension to the m-learning environment, namely �the Moodbile Project�.
Keywords:
Access control, e-learning, exam engine, Learning Management System (LMS), m- learning.
Abstract
Suraksha Digital Transformation to Indian Police Services
Meghana S, Hemashri R M, Aishwarya Nagesh Kamath
Abstract:
With the vision of Digital India programme to transform India into a digitally empowered society with safety of every individual is of highest priority in modern India. This project will be contributing to the Information and Communication Technology (ICT) which focuses on enabling effective communication between the citizen and the police. To increase transparency of Indian Police Services and to boost the public confidence rating, a mobile app that has all-in-one solution to enable secure real-time operational system will be developed as a part of the project. This project enables common man of India to get connected with the police system anytime, anywhere in India with various modules which includes Criminal booking module, Lost belongings module, Investigation module and Traffic monitoring module with accident rescue operations.
Keywords:
Suraksha, IPS, Mobile APP, Tracking ID, Digital Transformation, Police, User/Citizen.
Abstract
Video Recommendation
Ankit Agarwal, Kaveri M. Sultaniya, Mamatha.K.R
Abstract:
Today�s generation unlike the previous ones prefer the content with more multimedia and less of texts and as a result of which there is a enormous growth in services including multimedia and the huge amount of video contents that is offered in online social networks. Due to this huge variety and numbers available, the netizens have great trouble in getting their desired content. Therefore, a lot of personalized content recommendation systems have been suggested to fulfil the needs of all the individuals. However, the proposed systems tend to ignore the fact this social multimedia content data has led to the big data era and that�s where big data comes into picture, which has greatly obstructed the process of video recommendation. In addition, none of the proposed systems consider both the privacy of user data (e,g., personal details, social status, ages and hobbies) and video service provider database (where all the videos are stored), which are extremely sensitive and of very significant value both content wise and commercially. To handle these problems, we propose a video recommendation system based on cloud assistance and big data and also on distributed online learning. In our proposal, service vendors are mapped onto as distributed cooperative learners who recommends videos according to user�s context and liking, while simultaneously also adapts to the video-selection strategy which is based on number of user-clicks count to increase number of feedbacks or rewards.
Keywords:
recommender system; data privacy; big data; provider database;distributed learning.
Abstract
Wireless Body Area Network for Heart Attack Detection and Automatic Alert System
Pallavi Ganapati Gaonkar, Rishav Verma, Shashank K R
Abstract:
Heart assaults can happen at whatever time and anyplace. Utilizing unavoidable innovation, for example, a cell phone and a little wearable ECG sensor it is conceivable to gather the client's manifestations and to recognize the onset of a heart assault by examining the ECG recordings. On the off chance that the application surveys that the client is at hazard, it will ask the client to call the crisis benefits promptly. In the event that the client has a heart failure the application will naturally decide the present area of the client and caution the emergency vehicle administrations and others to the individual's area. The principle goal of this article is to create individual test heart assault application to identify potential casualties, without the intercession of restorative expert. The point is additionally to make the emergency vehicle benefit accessible at the patient's area, advised by the casualty's application in the wake of identifying heart assault. The application utilizes remote body region arrange for observing the heart heartbeats and distinguishing the heart assault. The BAN utilizes particular ECG sensors that recognizes human body's electrical properties, to empower surface-wave spread around the body.
Keywords:
ECG, Heart attack, Arduino.