The Random Character of Stock Market Prices.

Archives of disease in childhood. Fetal and neonatal edition

This paper defines software fairness and discrimination and develops a testing-based method for measuring if and how much software discriminates, focusing on causality in discriminatory behavior. Evidence of software discrimination has been found in modern software systems that recommend criminal sentences, grant access to financial products, and determine who is allowed to participate in promotions. Our approach, Themis, generates efficient test suites to measure discrimination. Given a schema describing valid system inputs, Themis generates discrimination tests automatically and does not require an oracle. We evaluate Themis on 20 software systems, 12 of which come from prior work with explicit focus on avoiding discrimination. We find that (1) Themis is effective at discovering software discrimination, (2) state-of-the-art techniques for removing discrimination from algorithms fail in many situations, at times discriminating against as much as 98% of an input subdomain, (3) Themis optimizations are effective at producing efficient test suites for measuring discrimination, and (4) Themis is more efficient on systems that exhibit more discrimination. We thus demonstrate that fairness testing is a critical aspect of the software development cycle in domains with possible discrimination and provide initial tools for measuring software discrimination.

https://dl.acm.org/doi/pdf/10.1145/3106237.3106277

Fairness testing: testing software for discrimination

Rather than tediously writing unit tests manually, tools can be used to generate them automatically --- sometimes even resulting in higher code coverage than manual testing. But how good are these tests at actually finding faults? To answer this question, we applied three state-of-the-art unit test generation tools for Java (Randoop, EvoSuite, and Agitar) to the 357 real faults in the Defects4J dataset and investigated how well the generated test suites perform at detecting these faults. Although the automatically generated test suites detected 55.7% of the faults overall, only 19.9% of all the individual test suites detected a fault. By studying the effectiveness and problems of the individual tools and the tests they generate, we derive insights to support the development of automated unit test generators that achieve a higher fault detection rate. These insights include 1) improving the obtained code coverage so that faulty statements are executed in the first instance, 2) improving the propagation of faulty program states to an observable output, coupled with the generation of more sensitive assertions, and 3) improving the simulation of the execution environment to detect faults that are dependent on external factors such as date and time.

/pdf/do-automatically-generated-unit-tests-find-real-faults-an-13f857kugj.pdf

Do Automatically Generated Unit Tests Find Real Faults? An Empirical Study of Effectiveness and Challenges (T)

Research on software testing produces many innovative automated techniques, but because software testing is by necessity incomplete and approximate, any new technique faces the challenge of an empirical assessment. In the past, we have demonstrated scientific advance in automated unit test generation with the EVOSUITE tool by evaluating it on manually selected open-source projects or examples that represent a particular problem addressed by the underlying technique. However, demonstrating scientific advance is not necessarily the same as demonstrating practical value; even if VOSUITE worked well on the software projects we selected for evaluation, it might not scale up to the complexity of real systems. Ideally, one would use large “real-world” software systems to minimize the threats to external validity when evaluating research tools. However, neither choosing such software systems nor applying research prototypes to them are trivial tasks. In this article we present the results of a large experiment in unit test generation using the VOSUITE tool on 100 randomly chosen open-source projects, the 10 most popular open-source projects according to the SourceForge Web site, seven industrial projects, and 11 automatically generated software projects. The study confirms that VOSUITE can achieve good levels of branch coverage (on average, 71p per class) in practice. However, the study also exemplifies how the choice of software systems for an empirical study can influence the results of the experiments, which can serve to inform researchers to make more conscious choices in the selection of software system subjects. Furthermore, our experiments demonstrate how practical limitations interfere with scientific advances, branch coverage on an unbiased sample is affected by predominant environmental dependencies. The surprisingly large effect of such practical engineering problems in unit testing will hopefully lead to a larger appreciation of work in this area, thus supporting transfer of knowledge from software testing research to practice.

/pdf/a-large-scale-evaluation-of-automated-unit-test-generation-24mhz66eca.pdf

A Large-Scale Evaluation of Automated Unit Test Generation Using EvoSuite

Test coverage is an important metric of software quality, since it indicates thoroughness of testing. In industry, test coverage is often measured as statement coverage. A fundamental problem of software testing is how to achieve higher statement coverage faster, and it is a difficult problem since it requires testers to cleverly find input data that can steer execution sooner toward sections of application code that contain more statements.We created a novel fully automatic approach for aChieving higher stAtement coveRage FASTer (CarFast), which we implemented and evaluated on twelve generated Java applications whose sizes range from 300 LOC to one million LOC. We compared CarFast with several popular test case generation techniques, including pure random, adaptive random, and Directed Automated Random Testing (DART). Our results indicate with strong statistical significance that when execution time is measured in terms of the number of runs of the application on different input test data, CarFast outperforms the evaluated competitive approaches on most subject applications.

/pdf/carfast-achieving-higher-statement-coverage-faster-3hehq9oi9i.pdf

CarFast: achieving higher statement coverage faster

Floods are one of the most frequently occurring disasters in Bangladesh that cause small to large scale damage every year. Most of the studies in the literature provide a flood damage prediction or inference model at individual building level. Some of the works that adopt a higher spatial scale such as households conduct their analysis on a few specific regions. This paper presents a household-level flood damage analysis performed on 2004–2009 flood data from 64 districts of Bangladesh. The study focuses both on prediction and determination of influencing factors because both of these facilitate flood damage reduction programs. A machine learning driven approach has been taken for prediction where three learning algorithms namely linear regression, random forest and artificial neural network are fitted to the data and compared. In this work, linear regression performed better than the other two because its assumptions were considered. A regression analysis showed the significance of the relationship between predictors and damage. Apart from the significant hydrologic predictors, literacy, flood awareness, house structure and disaster management knowledge were found to be influential. Preparedness was observed to be statistically insignificant unless it was combined with disaster management knowledge. A principal component analysis was further performed to cluster different variables into predictor groups and inspect their effect on flood damage. According to this analysis, hydrologic and environmental predictors, literacy, land ownership and house structure were found to be highly important where precaution and disaster related factors showed less significance.

A machine learning-based prediction and analysis of flood affected households: A case study of floods in Bangladesh

Many organizations deploy applications that use databases by sending Structured Query Language (SQL) statements to them and obtaining data that result from the execution of these statements. Since applications often share the same databases concurrently, database deadlocks routinely occur in these databases resulting in major performance degradation in these applications. Database engines do not prevent database deadlocks for the same reason that the schedulers of operating system kernels do not preempt processes in a way to avoid race conditions and deadlocks - it is not feasible to find an optimal context switching schedule quickly for multiple processes (and SQL statements), and the overhead of doing it is prohibitive. We created a novel approach that combines run-time monitoring, which automatically prevents database deadlocks, with static analysis, which detects hold-and-wait cycles that specify how resources (e.g., database tables) are held in contention during executions of SQL statements. We rigorously evaluated our approach. For a realistic case of over 1,200 SQL statements, our algorithm detects all hold-and-wait cycles in less than two seconds. We built a toolset and experimented with three applications. Our tool prevented all existing database deadlocks in these applications and increased their throughputs by up to three orders of magnitude.

https://www.cs.uic.edu/~drmark/index_htm_files/REDACT_FSE13.pdf

Preventing database deadlocks in applications

Despite lots of effort being exerted in designing feature descriptors, it is still challenging to find generalized feature descriptors, with acceptable discrimination ability, which are able to capture prominent features in various image processing applications. To address this issue, we propose a computationally feasible discriminative ternary census transform histogram (DTCTH) for image representation which uses dynamic thresholds to perceive the key properties of a feature descriptor. The code produced by DTCTH is more stable against intensity fluctuation, and it mainly captures the discriminative structural properties of an image by suppressing unnecessary background information. Thus, DTCTH becomes more generalized to be used in different applications with reasonable accuracies. To validate the generalizability of DTCTH, we have conducted rigorous experiments on five different applications considering nine benchmark datasets. The experimental results demonstrate that DTCTH performs as high as 28.08% better than the existing state-of-the-art feature descriptors such as GIST, SIFT, HOG, LBP, CLBP, OC-LBP, LGP, LTP, LAID, and CENTRIST.

/pdf/dtcth-a-discriminative-local-pattern-descriptor-for-image-25pwkb4fc7.pdf

DTCTH: a discriminative local pattern descriptor for image classification

Detecting spelling errors and correcting those errors automatically is a great research challenge. Developing a precise spell checker for Bangla language which detects spelling errors and provides suggestions for correcting those errors, is quite difficult because of the complex rules of Bangla spelling. In this paper, a clustering-based spell checking technique is proposed for Bangla language that reduces both search space and search time. Therefore, it improves the performance of a spell checker. The proposed spell checker can handle both typographical errors and phonetic errors. To evaluate the proposed spell checking technique, we use 2,450 misspelled words and the result shows that the proposed approach performs better for checking and correcting spelling errors. The success rate of proposed spell checker is 99.8%. We compare our spell checking technique with two Bangla spell checkers, Avro and Puspa and the proposed system provides relatively better results.

B M Mainul Hossain

Papers

CarFast: achieving higher statement coverage faster

A machine learning-based prediction and analysis of flood affected households: A case study of floods in Bangladesh

Preventing database deadlocks in applications

DTCTH: a discriminative local pattern descriptor for image classification

Clustering-based Bangla spell checker