Rinkle Rani

Modeling and querying data in NoSQL databases

Relational databases are providing storage for several decades now. However for todays interactive web and mobile applications the importance of flexibility and scalability in data model can not be over-stated. The term NoSQL broadly covers all non-relational databases that provide schema-less and scalable model. NoSQL databases which are also termed as Internetage databases are currently being used by Google, Amazon, Facebook and many other major organizations operating in the era of Web 2.0. Different classes of NoSQL databases namely key-value pair, document, column-oriented and graph databases enable programmers to model the data closer to the format as used in their application. In this paper, data modeling and query syntax of relational and some classes of NoSQL databases have been explained with the help of an case study of a news website like Slashdot.

Managing Data in Healthcare Information Systems: Many Models, One Solution

Because healthcare data comes from multiple, vastly different sources, databases must adopt a range of models to process and store it. A polyglot-persistent framework combines relational, graph, and document data models to accommodate information variety.

A parallel fuzzy clustering algorithm for large graphs using Pregel

We propose a parallel fuzzy clustering algorithm (PGFC) for large graphs.Pregel and Hadoop frameworks are used for processing of massive graph data.PGFC is scalable and produces good quality of clusters.The performance is validated using graphs having upto millions of nodes.Results reveal that PGFC significantly outperforms state-of-the-art algorithms. Large graphs are scale free and ubiquitous having irregular relationships. Clustering is used to find existent similar patterns in graphs and thus help in getting useful insights. In real-world, nodes may belong to more than one cluster thus, it is essential to analyze fuzzy cluster membership of nodes. Traditional centralized fuzzy clustering algorithms incur high communication cost and produce poor quality of clusters when used for large graphs. Thus, scalable solutions are obligatory to handle huge amount of data in less computational time with minimum disk access. In this paper, we proposed a parallel fuzzy clustering algorithm named PGFC for handling scalable graph data. It will be advantageous from the viewpoint of expert systems to develop a clustering algorithm that can assure scalability along with better quality of clusters for handling large graphs.The algorithm is parallelized using bulk synchronous parallel (BSP) based Pregel model. The cluster centers are initialized using degree centrality measure, resulting in lesser number of iterations. The performance of PGFC is compared with other state of art clustering algorithms using synthetic graphs and real world networks. The experimental results reveal that the proposed PGFC scales up linearly to handle large graphs and produces better quality of clusters when compared to other graph clustering counterparts.

IoT solutions for 3-D visualization of Twitter data

The Internet of Things (IoT) is a booming terminology that is used nowadays. It is interconnection of various objects such as RFID(radio frequency identifications), actuators, smart devices, sensors over an Internet. It is a interdisciplinary approach to study computing devices and their behavior. A lot of research is going on (IoT) devices to gain the intrinsic potential from them. Over the last few years, we have seen great efforts from industry and academia to provide IoT solutions and develop customer oriented market for IOT devices. In this paper, we will examine the various IoT solutions for twitter data visualization and analysis. Twitter being a popular social networking site always gives researchers a scope for innovation. A wide variety of solutions have been proposed for 3-D visualization of twitter data, but “Twitter Mood Light” is popular among all.

Classification of Cancerous Profiles Using Machine Learning

There are a variety of options available for cancer treatment. The type of treatment recommended for an individual is influenced by various factors such as cancer-type, the severity of a cancer (stage) and most important the genetic heterogeneity. In such a complex environment, the targeted drug treatments are likely to be irresponsive or respond differently. To study anti-cancer drug response we need to understand cancerous profiles. These cancerous profiles carry information which can reveal the underlying factors responsible for cancer growth. Hence, there is need to analyze cancer data for predicting optimal treatment options. Analysis of such profiles can help to predict and discover potential drug targets and drugs. In this paper the main aim is to provide machine learning based classification technique for cancerous profiles.

A comparative study of elasticsearch and CouchDB document oriented databases

With the advent of large complex datasets, NOSQL databases have gained immense popularity for their efficiency to handle such datasets in comparison to relational databases. There are a number of NOSQL data stores for e.g. Mongo DB, Apache Couch DB etc. Operations in these data stores are executed quickly. In this paper we aim to get familiar with 2 most popular NoSQL databases: Elasticsearch and Apache CouchDB. This paper also aims to analyze the performance of Elasticsearch and CouchDB on image data sets. This analysis is based on the results carried out by instantiate, read, update and delete operations on both document-oriented stores and thus justifying how CouchDB is more efficient than Elasticsearch during insertion, updation and deletion operations but during selection operation Elasticsearch performs much better than CouchDB. The implementation has been done on LINUX platform.

Comparative analysis of density based outlier detection techniques on breast cancer data using hadoop and map reduce

Advancement of technology, has furnished several terabytes of data for companies which can be effectively summed under Data Mining. Finding useful pieces of information from such huge data has been the need of the hour. A term called Anomaly Detection [8] is used in the pretext to refer to data objects which do not confer to a notion of normal data objects. There are various density based clustering algorithms[10] used to categorize data objects as normal or anomalous by finding clusters within the data set. LOF[18] finds the anomalous data objects by finding local density of data objects with respect to local density of its neighbors. DBSCAN finds anomalous data objects by finding data objects surrounded by data objects (density) which are far away from the concerned data object. OPTICS an extension of DBSCAN finds clusters of arbitrary sizes. DENCLUE uses a set of density distribution functions. This paper shows the comparison of the density based algorithms i.e. LOF, OPTICS, DBSCAN, DENCLUE based upon parameters such as time taken on single cluster hadoop, noise accuracy detection level, number of anomalous instances detected on high dimensional data, handle varied density, input parameters and complexity etc.

A 3-level model for implementing MOOC in India

In the recent years the concept of distant learning and online courses has gained great popularity. MOOC (Massive Open Online Course) is the new trend on the Internet and has rapidly gained much popularity. MOOC allows learners from all over the world to learn in a connected way and in unprecedented scale. MOOC can prove to be a great helping hand in a country like India with more than five thousand engineering colleges [1]. These are gaining huge attention from students all over the globe. These MOOC are acting as a revolution in the field of education. Collective learning and massive involvement of individuals are the key features that make MOOC impactful. Offering of such courses by elite universities lure the individuals to get registered for these courses. This paper aims to propose a model for Indian education system using BLMM [15] as a base for the improvement. Thus it will first describe the general scenario of MOOCs in India and later will propose three level implementation models for Indian education system.

Ap-FSM: A Parallel Algorithm for Approximate Frequent Subgraph Mining using Pregel

Abstract Large graphs are scale-free, ubiquitous having irregular relationships and non-trivial topology. Frequent subgraph mining is a popular method for knowledge extraction from graphs. Most of the existing frequent subgraph mining algorithms are centralized algorithms that cannot handle a single large graph efficiently and incur high communication cost. However, to make the task of subgraph mining less expensive computationally, approximate subgraph mining can be applied which will capture similar structure subgraphs as of exact subgraph mining. In this paper, we propose an approximate subgraph mining algorithm named Ap-FSM implemented on distributed graph environment Pregel. The working of Ap-FSM is divided into three phases. The first phase selects the representative graph from the original graph while preserving the original graph properties. The second phase efficiently performs subgraph extension. Phase 3 introduces a novel two-step optimization for performing subgraph pruning. Analyzing such large graph data will be beneficial from the perspective of expert and intelligent systems, as discovered patterns can be used for knowledge discovery and decision making. To evaluate the performance of Ap-FSM, experiments are performed over three real life datasets having up to billion edges. The results show that the proposed Ap-FSM significantly outperforms the state-of-art frequent subgraph mining algorithms and overcome the challenges of performing frequent subgraph mining on a massive large graph. It is also shown that Ap-FSM achieves high scalability and speedup in distributed graph environment and is highly accurate in finding frequent subgraphs from a single large graph.

Systematic Review of Clustering High-Dimensional and Large Datasets

Technological advancement has enabled us to store and process huge amount of data in relatively short spans of time. The nature of data is rapidly changing, particularly its dimensionality is more commonly multi- and high-dimensional. There is an immediate need to expand our focus to include analysis of high-dimensional and large datasets. Data analysis is becoming a mammoth task, due to incremental increase in data volume and complexity in terms of heterogony of data. It is due to this dynamic computing environment that the existing techniques either need to be modified or discarded to handle new data in multiple high-dimensions. Data clustering is a tool that is used in many disciplines, including data mining, so that meaningful knowledge can be extracted from seemingly unstructured data. The aim of this article is to understand the problem of clustering and various approaches addressing this problem. This article discusses the process of clustering from both microviews (data treating) and macroviews (overall clustering process). Different distance and similarity measures, which form the cornerstone of effective data clustering, are also identified. Further, an in-depth analysis of different clustering approaches focused on data mining, dealing with large-scale datasets is given. These approaches are comprehensively compared to bring out a clear differentiation among them. This article also surveys the problem of high-dimensional data and the existing approaches, that makes it more relevant. It also explores the latest trends in cluster analysis, and the real-life applications of this concept. This survey is exhaustive as it tries to cover all the aspects of clustering in the field of data mining.