Computer Science and Engineering

<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

The explosive growth of social network with the help of e-commerce websites has made the issue of information retrieval increasingly challenging. Users may not have the time or knowledge to personally evaluate the options which are available on social network platform. Recommender systems present themselves as a practical answer to endless options available online. In this research paper, we give an insight into various types of filtering techniques associated with recommender systems. We also discuss the problems faced by these filtering techniques and evaluation criteria on the basis of which various algorithms made for recommendation purpose can be compared. Finally we propose a composite news recommendation system model. © 2015 IEEE.

World University root node

(SoE) School of Engineering

2015 4th International Conference on Reliability, Infocom Technologies and Optimization: Trends and Future Directions, ICRITO 2015

Evaluation criteria for measuring the performance of recommender systems

Link prediction problem may be applied in many different domains. In the network, nodes represent authors and a connection between them indicates they coauthored a paper. This paper describes an improved method for predicting possibility of a link between authors. The algorithm is implemented in C# programming language and testing has been performed on Intel Core2Duo processor with 4 GB RAM and 2 GHz processor speed. It was observed that this method shows better performance compared to the previously presented methods. © 2014 IEEE.

Proceedings - 2014 3rd International Conference on Reliability, Infocom Technologies and Optimization: Trends and Future Directions, ICRITO 2014

Link prediction algorithm for co-authorship networks using Neural Network

The proposed method addresses human action recognition problem in a realistic video. The content of such videos are influenced by irregular background motion and camera shakes. We construct the human pose descriptors by using a modified version of optical flow (we call it as hybrid motion optical flow). We quantize the hybrid motion optical flow (HMOF) into different labels. The orientations of the HMOF vectors are corrected using probabilistic relaxation labelling, where the HMOF vectors with locally maximum magnitude are retained. A sequence of 2D points, called tracks, representing the motion of the person, are constructed. We select top dominant tracks of the sequence based on a cost function. The dominant tracks are further processed to represent the feature descriptor of a given action. © Springer International Publishing Switzerland 2015.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Human action recognition using dominant motion pattern

This paper presents a dynamic model that can provide prediction for the estimated arrival time of a bus at a given bus stop using Global Positioning System (GPS) data. The proposed model is a hybrid intelligent system combining Fuzzy Logic and Neural Networks. While Neural Networks are good at recognizing patterns and predicting, they are not good at explaining how they decide their input parameters. Fuzzy Logic systems, on the other hand, can reason with imprecise information, but require linguistic rules to explain their fuzzy outputs. Thus combining both helps in countering each other’s limitations and a reliable and effective prediction system can be developed. Experiments are performed on a real-world dataset and show that our method is effective in stated conditions. The accuracy of result is 86.293% obtained. © Springer International Publishing Switzerland 2015.

Bus arrival time prediction using a modified amalgamation of fuzzy clustering and neural network on spatio-temporal data

Social networking applications are changing the way people communicate with their peers and family. Social networking applications have several features that help in connecting one to another, from finding old friends to browsing for new friends and contacts. To harness the true power of social networking, sometimes it may be important to seamlessly identify and connect individuals who are geographically nearby with similar interests at a given time. In this work, we develop and discuss a pervasive and context-aware social networking system in a mobile environment that is able to report the presence of like-minded individuals in real time based on user's location. We discuss how our system computes closeness between two individuals and develop an application to test its efficacy. © 2014 IEEE.

2014 IEEE International Conference on Advanced Networks and Telecommunication Systems, ANTS 2014

PECSNET: The next generation pervasive and context aware social network

Mobile agents in wireless sensor networks provide many advantages in comparison of conventional client/server architecture. Mobile Agents follow code to data approach and thus perform the task of data aggregation at the nodes rather than at the processing element. This approach reduces in-network data transmission, thus improving bandwidth usage and prolonging network lifetime. Use of multiple mobile agents instead of single mobile agent allows the task to get completed in a cooperative manner. In this paper, we present a multi agent solution for data aggregation using source grouping along with tree based ordering. Extensive simulations have proved better results in terms of energy consumed and latency in comparison of other approaches. © 2014 IEEE.

Proceedings of the 2014 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2014

Multi-Agent data aggregation in wireless sensor network using source grouping

There has been continuous development in the wildlife DNA forensics research that relied on the collection and analysis of the biological samples over the past many years. But there is not enough progress to develop computational algorithms which could make the process of finding the origin of species easier and faster. Computational algorithms based on phylogenetic networks are capable of providing evidence to assist in wildlife law enforcement and species conversation. Our new algorithmic rule first rudiment to spot sample set, provides a promising new model for the strong reasoning of substructure and ancestry of wildlife trade. Our findings establish a clear evolutionary connection among many different problem sets.

ACM International Conference Proceeding Series

Detecting history of species using mining of motifs in phylogenetic networks

A comparative study of the various motif search algorithms is very important for several reasons. For example, we could identify the strengths and weaknesses of each. As a result, we might be able to devise hybrids that will perform better than the individual components. In this paper, we (either directly or indirectly) compare the performance of PMSprune (an algorithm based on the (l, d)-motif model) and several other algorithms in terms of seven measures and using well-established benchmarks. We have employed several benchmark datasets including the one used by Tompa et al. It is observed that both PMSprune and DME (an algorithm based on position-specific score matrices), in general, perform better than the 13 algorithms reported in Tompa et al. Subsequently, we have compared PMSprune and DME on other benchmark datasets including ChIP-Chip, ChIP-Seq and ABS. Between PMSprune and DME, PMSprune performs better than DME on six measures. DME performs better than PMSprune on one measure (namely, specificity). Copyright © 2014 Inderscience Enterprises Ltd.

Fulltext

International Journal of Bioinformatics Research and Applications

An experimental comparison of PMSprune and other algorithms for motif search

With the wide spread usage of LBS, convenience has reached on the finger tips of mobile users, but on the other side, it has escalated many security and privacy concerns. In this paper we address the location K-anonymity problem using fuzzy spatio-temporal attributes, a new perspective of looking at privacy issue in location privacy. In the context of LBSs and mobile clients, location K-anonymity refers to K-anonymous usage of location information. A novel approach for determining location disclosure based on fuzzy attributes of spatio-temporal context is proposed which in turn will give us a value of K for K-anonymity purpose. Spatio-temporal fuzzy attributes for privacy issues are identified and Fuzzy Inference System (FIS) is implemented that takes these attributes as input and generates location disclosure value. Using Location disclosure value, K is computed for K-anonymity to ensure privacy. This value of K is directly based on current spatio temporal context and is valid for all users present in that context. Further, an exhaustive rule base of fuzzy rules is generated based on responses obtained by conducting survey on the potential users who frequently use POI (Point of Interest) services. Later on, fuzzy rules for FIS rule base are extracted using Fuzzy C Means (FCM) clustering technique. Using the rules extracted through FCM, the size of rule base is reduced and the performance of the FIS is evaluated. Number of rules in rule base is decreased for scalability and efficiency purposes. Root Mean Square Error (RMSE) for every reduced set is computed and compared with initial exhaustive rule base. It is observed that size of rule base can be decreased to a considerable extent. © 2014 IEEE.

Proceedings - IEEE International Conference on Mobile Data Management

K anonymity based on fuzzy spatio-temporal context

With the increase in number of available interesting locations, it becomes difficult for users to find interesting ones, thus imposes a need for recommender systems to suggest interesting locations. Further, to ease the user's decision making, the amount of supplementary information, such as right time to visit, weather conditions, traffic condition, right mode of transport, crowdedness, security alerts, etc., may be annotated with the list of recommended locations. This paper explores the possibility of enriching tourist locations using crowd sourcing approach, which can be used by Tourist Spot Recommender System (TSRS) for mobile users. Proposed crowd sourcing system focuses on getting work done from the crowd currently available at the location under consideration. In proposed system, the contributed information is not limited to ones available on blogs, web pages and sensor-readings from the device etc., but includes proactively-generated user's opinions and perspectives, that are processed to offer immediate knowledge. Our system works in collaboration with a TSRS, takes the list of locations to be recommended to the current user and performs just-in-time information enrichment for those selected set of locations. We have implemented a prototype of proposed systems using java android software development toolkit and evaluated this system by 76 real users. © 2014 IEEE.