Search

Prediction of student performance at early stage in higher education is important for academic society so that strategic decisions can be made before students are placed to keep them from dropping out of the course. Due to India's massive student population and extremely ancient educational system, there are significant difficulties in measuring and forecasting students' performance. Every institution in India has its own unique set of criteria for measuring student achievement, and there is no formal process for keeping track of and evaluating a student's progress and improvement. Over the last decade, researchers in the education domain have presented numerous types of machine learning techniques. However, there are significant obstacles to dealing with imbalanced datasets in order to predict the performance of students. In this paper, the first phase of traditional classification algorithms has been applied to the dataset, which contains the progress of 4424 students. In the second phase, novel hybrid machine learning (ML) algorithms were used to get better predictions. The outcome of the proposed model makes it easier to predict how well students will do so that early decisions can be made about the growth of higher education institutions.

Nowadays, complex network has been inspired by the observation of many real-world network patterns such as pattern of mobile phone usage, gene expression, e-mail habits, protein interactions in networks, regional interaction in archaeological contexts and co-authorship in research publications. The detection of interaction patterns that occur at fixed time periods can be mined in a complex network. This article proposes an algorithm to predict an interaction pattern hidden behind these networks, namely periodic subgraphs embedded in a complex network. The primary data structures used in tracking the periodic subgraphs are lists, tables, data frames, matrices and vectors. The algorithm can be implemented using analytical tools that do not require advanced data structures such as pattern trees and indexing techniques such as hash maps. In this paper, our algorithm processes a dynamic network in two stages. The first stage of the algorithm extracts all maximal subgraphs which appear between successive time steps. In second stage, it maintains the periodicities of the maximal subgraphs identified and flushes out those graphs whose periodicities ceased to exist. The algorithm is tested by generating a dynamic network using random graph generating function and by superimposing periodic subgraphs on the network at different time interval periods.

Objectives: To analyze the issue of cold-start (user cold-start and item cold-start) in Collaborative Filtering Recommender System (CFRS) and to compare its solution with various approaches are summarized in this paper. Methods/Statistical Analysis: The manuscript discussed about the cold-start issue in which the recommender system cannot recommend items to the new user because no ratings made by the new user (user cold-start) as well as for the newly added items, the system cannot be able to provide recommendations to the user because the system has no ratings for the newly added item (item cold-start). The solutions for cold-start issue are analyzed based on the model based approach, demographic data, ask-to-rate technique, and Social Network Analysis (SNA). Findings: The comparative review of the aforementioned approaches provides the detail about how to implement the model based approach, how to collect the demographic data from the new user, how to apply the ask-to-rate technique and how to make use of the SNA concept to solve the cold-start issue in CF recommender system. Application: The recommender system on Amazon helps the user to purchase books, Compact Disks (CDs), Netflix helps the user to choose CDs to purchase/rent and Epinions, helps the users to decide to purchase based on user reviews. Keywords: Ask-To-Rate Technique, Cold-Start, Collaborative Filtering Recommender System, Demographic Data, Model Based Approach, Social Network Analysis

Coffee is the natural gift of Ethiopia. Generally the export quality washed coffee beans of Ethiopia are classified into two grades and sundried coffee beans into five different grades based on their number of defects. The objective of the research is to extract the features of green coffee beans from the images which would be helpful on classifying the coffee beans to different grades by an automated system. Different image processing techniques are applied on the images to perform preprocessing, segmentation and feature extraction. The extracted features of the coffee beans from images are broadly classified as morphological, textural and color. The morphological feature includes area, perimeter, major axis length, minor axis length, Eccentricity. Energy, Entropy, contrast and homogeneity are the information relevant to texture. Individual color component values of the three primary colors, along with its hue, intensity and saturation values are extracted for color features of the coffee beans. Automated classification and machine learning algorithms needs a data set for further processing. Data mining applications for discovering different patterns for various grades and knowledge discovery need a highly accurate dataset. Hence, preparing such data set by applying image processing techniques is becoming the objective of this research study. The objective is realized with a dataset of 100 observations for each grade with morphological, textural, and color features.

Recommender system (RS) has become a very important factor in many eCommerce sites. In our daily life, we rely on the recommendation from other persons either by word of mouth, recommendation letters, movie, item and book reviews printed in newspapers, etc. The typical Recommender Systems are software tools and techniques that provide support to people by identifying interesting products and services in online store. It also provides a recommendation for certain users who search for the recommendations. The most important open challenge in Collaborative filtering recommender system is the cold start problem. If the adequate or sufficient information is not available for a new item or users, the recommender system runs into the cold start problem. To increase the usefulness of collaborative recommender systems, it could be desirable to eliminate the challenge such as cold start problem. Revealing the community structures is crucial to understand and more important with the increasing popularity of online social networks. The community detection is a key issue in social network analysis in which nodes of the communities are tightly connected each other and loosely connected between other communities. Many algorithms like Givan-Newman algorithm, modularity maximization, leading eigenvector, walk trap, etc., are used to detect the communities in the networks. To test the community division is meaningful we define a quality function called modularity. Modularity is that the links within a community are higher than the expected links in those communities. In this paper, we try to give a solution to the cold-start problem based on community detection algorithm that extracts the community from the social networks and identifies the similar users on that network. Hence, within the proposed work several intrinsic details are taken as a rule of thumb to boost the results higher. Moreover, the simulation experiment was taken to solve the cold start problem.

In a given geometric objects, set of objects are retrieved as results for the given query. The problem is defined as geometric query intersection problem. The problem defines that, for the given user query, all the relevant results which are intersected by the query are showed as output. Because of the large number results, it occupies the huge memory space. For that here we propose a solution that we show the top-k results as the output where k is the integer. This paper gives a general technique to solve any top-k GIQ problem efficiently. The technique relies only on the availability of an efficient solution for the underlying (non-top-k) GIQ problem, which is often the case. Using this, asymptotically efficient solutions are derived for several top-k GIQ problems, including top-k orthogonal and circular range search, point enclosure search, half space range search, etc. Implementations of some of these solutions, using practical data structures, show that they are quite efficient in practice. This paper also does a formal investigation of the hardness of the top-k GIQ problem, which reveals interesting connections between the top-k GIQ problem and the underlying (non-top-k) GIQ problem. In our proposed system we are going to perform multiple query processing. Because in the existing approach it can able to process only one query at the time. But in our proposed work we process multiple queries at the time.

The secure social networks are online based services that permit unique users to build a public profile, to build the number of users with whom they want to communicate and see the connections within the system. Many social networks are online based and it gives the users to communicate via the internet, such as electronic mailing, quick messaging, sharing photos and videos, uploading the thoughts of the users etc. Such social networks are easily affected by attackers (i.e Sybil attack). The Sybil attack is a kind of security threat and it cause when an insecure system is hijacked to claim various identities. The huge scale peer to peer systems meets security thread from damaged calculating fundamentals. The Robust Recommendation algorithm is used to overcome the Sybil attack that affects the application but it fails when the attacker knows about the profiles of authenticate users. The proposed methodology has to eliminate such constraints by using the combination of session management and face detection and recognition techniques. By using these procedures, the application is secured effectively.

Confidential information exchange electronically in a secure way becomes very important in the age of computers. A newly proposed public key crypto method based on Chebyshev polynomials [16] recommend a new approach to data encryption. This paper provides an overview of three major types of public key cryptosystems mainly Diffie-Hellman Key Exchange Algorithm, the RSA Cryptosystem, the ElGamal Encryption Method and their implementation using Chebyshev Polynomials. During the past decade, there has been tremendous interest world-wide image encryption which is based on edge detail. In this method the image divided into two, one is the image contains edges and another is the image without edges, and encrypt the both images using encryption algorithm. The encrypted images are complex for intruder to decode, so it provides a high level of security.