An Approach to Security for Unstructured Big Data

Security of Big Data is a huge concern. In a broad sense, Big Data contains two types of data: structured and unstructured. Providing security to unstructured data is more difficult than providing security to structured data. In this paper, we have developed an approach to provide adequate security to unstructured data by considering types of data and their sensitivity levels. We have reviewed the different analytics methods of Big Data to build nodes of different types of data. Each type of data has been classified to provide adequate security and enhance the overhead of the security system. To provide security to a data node, and a security suite has been designed by incorporating different security algorithms. Those security algorithms collectively form a security suite which has been interfaced with the data node. Information on data sensitivity has been collected through a survey. We have shown through several experiments on multiple computer systems with varied configurations that data classification with respect to sensitivity levels enhances the performance of the system. The experimental results show how and in what amount the designed security suite reduces overhead and increases security simultaneously.     

Contextual Innovation Management Using a Stage‐Gate Platform: The Case of Philips Shaving and Beauty

To improve its innovation process, Philips Shaving and Beauty (S&B) designed a blueprint for its innovation process. Although it has proved to be quite effective, it has experienced a lack of efficiency, in terms of frequent cost and time overruns, in the fuzzy front end of this process. We suggest a contextual innovation management approach to set up a stage‐gate‐based innovation process platform and thus improve the efficiency in the fuzzy front end, which means that, for different contexts, stage‐gate process variants will be designed from which unnecessary activities are removed and important activities are emphasized. The design is based on the identification of relevant contextual factors to develop variations of the common innovation process within Philips S&B. We distinguished different variants of the innovation processes within Philips S&B that can increase the efficiency in the fuzzy front end. Based on interviews within and outside Philips S&B, we identified problems and potential solutions with regard to efficiency in eight recently finished innovation processes. The results indicate that the most important contextual factors are the distinction between incremental and radical innovations, and between market and technology‐based innovations. We used these factors to design three variants on the basic platform of the stage‐gate process.     

Migrant Remittances Provide Resilience Against Disasters in Africa

How responsive are migrant remittances to various disasters, both natural and human-made? Would remittances be affected by systemic financial crises, such as the 2008–09 financial crisis, or more recent crises affecting the Eurozone? Using panel data on 23 sub-Saharan African (SSA) countries from 1980 to 2007, we find that remittances are slow to respond to natural disasters, unresponsive to outbreaks of conflict, and will slowly decline following a systemic financial crisis. This suggests that, given its stability, remittances are sources of resilience in SSA.     

Visual displays: analytical study and applications to graphs and real data

Principal axis methods such as principal component analysis (PCA) and correspondence analysis (CA) are useful for identifying structures in data through interesting planar graphic displays. However, some kinds of data sets can be dealt alternatively with PCA or CA. This paper focuses on methods, such as PCA and CA, and on visual displays. Our aim is to illustrate the implications for a potential user of selecting either method, and its advantages and disadvantages, from an applied point of view. This is a matter covered broadly in textbooks and elsewhere considering theoretical arguments. Our purpose is to contribute to the comparison between these methods, over the same data set, in order to illustrate them for the practitioner. In the first part of this paper we present a novel analytical study of a binary matrix associated with a non-oriented axis-symmetric graph and show that CA outperforms standardized PCA for the reconstitution and visualization of such kind of graphs. In the second part we present a case using real data dealing with the distribution of employees in different economic sectors for the countries of the European Union, analyzed by means of standardized PCA and two-way CA, in order to see the differences between the two methods in practice.     

THE MEANING OF FAILED REPLICATIONS: A REVIEW AND PROPOSAL

The welcome rise of replication tests in economics has not been accompanied by a consensus standard for determining what constitutes a replication. A discrepant replication, in current usage of the term, can signal anything from an unremarkable disagreement over methods to scientific incompetence or misconduct. This paper proposes a standard for classifying one study as a replication of some other study. It is a standard that places the burden of proof on a study to demonstrate that it should have obtained identical results to the original, a conservative standard that is already used implicitly by many researchers. It contrasts this standard with decades of unsuccessful attempts to harmonize terminology, and argues that many prominent results described as replication tests should not be described as such. Adopting a conservative standard like this one can improve incentives for researchers, encouraging more and better replication tests.