Negative results for software effort estimation

More than half the literature on software effort estimation (SEE) focuses on comparisons of new estimation methods. Surprisingly, there are no studies comparing state of the art latest methods with decades-old approaches. Accordingly, this paper takes five steps to check if new SEE methods generated better estimates than older methods. Firstly, collect effort estimation methods ranging from “classical” COCOMO (parametric estimation over a pre-determined set of attributes) to “modern” (reasoning via analogy using spectral-based clustering plus instance and feature selection, and a recent “baseline method” proposed in ACM Transactions on Software Engineering). Secondly, catalog the list of objections that lead to the development of post-COCOMO estimation methods. Thirdly, characterize each of those objections as a comparison between newer and older estimation methods. Fourthly, using four COCOMO-style data sets (from 1991, 2000, 2005, 2010) and run those comparisons experiments. Fifthly, compare the performance of the different estimators using a Scott-Knott procedure using (i) the A12 effect size to rule out “small” differences and (ii) a 99 % confident bootstrap procedure to check for statistically different groupings of treatments. The major negative result of this paper is that for the COCOMO data sets, nothing we studied did any better than Boehms original procedure. Hence, we conclude that when COCOMO-style attributes are available, we strongly recommend (i) using that data and (ii) use COCOMO to generate predictions. We say this since the experiments of this paper show that, at least for effort estimation, how data is collected is more important than what learner is applied to that data.     

Beyond utilitarian factors: User experience and travel company website successes

Travel websites are extensively used by travelers for their travel bookings, yet little is known about the hedonic effectiveness of such experiences for the customer. The end users of these websites are customers who may not always be sophisticated information system (IS) users. Further, traditional utility based measures of customer evaluation are dated with respect to the interactive nature of the technology in use and the hedonic benefits that may result from the use of the technology. The evaluation of IS needs to include hedonic measures facilitated by the interactive technology in addition to the traditional utility based measures. We propose and test an evaluation model for retail travel websites that combines the traditional utility based measures with hedonic measures which collectively create a more comprehensive measure for the IS evaluation of consumer focused websites. Thus rooted in theory, the model extends DeLone and McLean’s model of IS Success by adding the construct of User Experience. The model was tested on a sample of 255 customers of travel websites targeted at the Indian market. The results indicate that both utility based and hedonic measures are important factors for customer IS (travel website in this study) evaluation. The larger implications for theory and practice of IS evaluation are explained.     

Using RFID for the management of pharmaceutical inventory — system optimization and shrinkage control

Motivated by a case study at a radiology practice, we analyze the incremental benefits of RFID technology over barcodes for managing pharmaceutical inventories. Unlike barcode technology, RFID enables accurate real-time visibility, which in turn enables several process improvements. We analyze the impact of automatic counting and discuss the system redesign critical to optimizing the inventory policy and eliminating shrinkage. We show that continuous review is superior to periodic review whenever accurate real-time information is available at no additional cost. We explain how RFID-enabled strategies vary with inventory parameters and provide a cost-benefit analysis for the implementation of RFID for the radiology practice.     

Fast and versatile texture-based wireframe rendering

This paper revisits the problem of wireframe rendering, which, at first, appears to be an easily solved problem. However, the conventional solution is inefficient and does not result in high-quality images. Recently, graphics hardware programming has been employed to achieve high-quality solid line rendering. In this paper, we present a simpler and faster technique for wireframe rendering based on texture mapping. Our technique does not require (but can benefit from) graphics hardware programming and thus can be easily integrated into existing rendering engines, while resulting in fast, accurate, high-quality, anti-aliased, and still versatile, wireframe drawing. For topologically structured meshes, our approach allows the rendering of wireframe decoupled from the underlying mesh, making possible the rendering of original wireframes on top of decimated meshes.     

Structural, electronic, and magnetic properties of heterofullerene C(58)Si with odd number of atoms and a near planar tetracoordinate Si atom

Density functional calculations and minimization techniques have been employed to characterize the structural and electronic properties of [5,6]-heterofullerene-C₅₈Si-C_2v. Since it has odd number of atoms and a near planar tetracoordinate Si atom on the skeleton of the cage, it has odd number of atoms assembling a cage and is a novel molecule. Vibrational frequencies of the molecule have been calculated at the B3LYP/6-31G* level of theory. The absence of imaginary vibrational frequency confirms that the molecule corresponds to a true minimum on the potential energy hypersurface. Sixteen ¹³C nuclear magnetic resonance (NMR) spectral signals of C₅₈Si are characterized, and its heat of formation was estimated in this work.     

Automatic Clustering Using an Improved Differential Evolution Algorithm

Differential evolution (DE) has emerged as one of the fast, robust, and efficient global search heuristics of current interest. This paper describes an application of DE to the automatic clustering of large unlabeled data sets. In contrast to most of the existing clustering techniques, the proposed algorithm requires no prior knowledge of the data to be classified. Rather, it determines the optimal number of partitions of the data "on the run." Superiority of the new method is demonstrated by comparing it with two recently developed partitional clustering techniques and one popular hierarchical clustering algorithm. The partitional clustering algorithms are based on two powerful well-known optimization algorithms, namely the genetic algorithm and the particle swarm optimization. An interesting real-world application of the proposed method to automatic segmentation of images is also reported.     

A parametric bootstrap approach for ANOVA with unequal variances: Fixed and random models

This article is about testing the equality of several normal means when the variances are unknown and arbitrary, i.e., the set up of the one-way ANOVA. Even though several tests are available in the literature, none of them perform well in terms of Type I error probability under various sample size and parameter combinations. In fact, Type I errors can be highly inflated for some of the commonly used tests; a serious issue that appears to have been overlooked. We propose a parametric bootstrap (PB) approach and compare it with three existing location-scale invariant tests—the Welch test, the James test and the generalized F (GF) test. The Type I error rates and powers of the tests are evaluated using Monte Carlo simulation. Our studies show that the PB test is the best among the four tests with respect to Type I error rates. The PB test performs very satisfactorily even for small samples while the Welch test and the GF test exhibit poor Type I error properties when the sample sizes are small and/or the number of means to be compared is moderate to large. The James test performs better than the Welch test and the GF test. It is also noted that the same tests can be used to test the significance of the random effect variance component in a one-way random model under unequal error variances. Such models are widely used to analyze data from inter-laboratory studies. The methods are illustrated using some examples.     

Discrete choice in commuter space: Small area analysis of diurnal population change in the Tokyo Metropolitan Region

Considerations of metropolitan commuting have often related to urban space as a continuum, and to commuting itself as a process. This approach has led to escalating complexity in the measurement, modeling and analysis of commuting, often with no commensurate results. The present study extends an alternative approach that considers commuter space of a metropolis as a discrete set of small areas. Each small area throughout the metropolis is a nighttime location of some households and, as a possible discrete destination choice of individuals within households, each subarea is also a daytime location of some persons. Each subarea, therefore, may be thought of as attaining two values, corresponding to its daytime (e.g., noontime) and nighttime (e.g., midnight) populations. The diurnal demographic shift of subareas, as a set of binary events, or toggles, between daytime and nighttime populations, is associated with commuter access throughout the metropolis. Diurnal shift offers structural simplicity aimed at estimating small-area daytime populations across a metropolitan region that might suitably precede and complement conventional continuum considerations of commuting. Using data for the Tokyo Metropolitan Region, we identify average household size as the nighttime indicator of small-areas, and daytime population density as their daytime indicator. The potential for the estimation of daytime populations across metropolitan regions where daytime population data are unavailable, is underscored by the clear tendency towards inverse relation of these two indicators of diurnal shift.     

Contrary to time conditionals in Talmudic logic

We consider conditionals of the form A ? B where A depends on the future and B on the present and past. We examine models for such conditional arising in Talmudic legal cases. We call such conditionals contrary to time conditionals. Three main aspects will be investigated:

1. Inverse causality from future to past, where a future condition can influence a legal event in the past (this is a man made causality).
2. Comparison with similar features in modern law.
3. New types of temporal logics arising from modelling the Talmudic examples.

We shall see that we need a new temporal logic,which we call Talmudic temporal logic with linear open advancing future and parallel changing past, based on two parameters for time.