Abstract—[Context]: The numerous challenges that can hin-der software companies from gathering their own data have motivated over the past 15 years research on the use of cross-company (CC) datasets for software effort prediction. Part of this research focused on Web effort prediction, given the large increase worldwide in the development of Web applications. Some of these studies indicate that it may be possible to achieve better performance using CC models if some strategy to make the CC data more similar to the within-company (WC) data is adopted. [Goal]: This study investigates the use of a recently proposed approach called Dycom to assess to what extent Web effort predictions obtained using CC datasets are effective in relation to the predictions obtained using WC data when explicitly mapping the CC models to the WC context. [Method]: Data on 125 Web projects from eight different companies part of the Tukutuku database were used to build prediction models. We benchmarked these models against baseline models (mean and median effort) and a WC base learner that does not benefit of the mapping. We also compared Dycom against a competitive CC approach from the literature (NN-filtering). We report a company-by-company analysis. [Results]: Dycom usually managed to achieve similar or better performance than a WC model while using only half of the WC training data. These results are also an improvement over previous studies that investigated the use of different strategies to adapt CC models to the WC data for Web effort estimation. [Conclusions]: We conclude that the use of Dycom for Web effort prediction is quite promising and in general supports previous results when applying Dycom to conventional software datasets. I.

Abstract. Software companies exploit data about completed projects to estimate the development effort required for new projects. Software size is one of the most important information used to this end. However, different methods for sizing software exist and companies may require to migrate to a new method at a certain point. In this case, in order to exploit historical data they need to resize the past projects with the new method. Besides to be expensive, resizing is also often not possible due to the lack of adequate documentation. To support size measurement migration, we propose a transfer learning approach that allows to avoid resizing and is able to estimate the effort of new projects based on the combined use of data about past projects measured with the previous measurement method and projects measured with the new one. To assess our proposal, an empirical analysis is carried out using an industrial dataset of 25 projects. Function Point Analysis and COSMIC are the measurement methods taken into account in the study.

This paper formulates app store analysis as an instance of software repository mining. We use data mining to extract feature information, together with more readily available price and popularity information, to support analysis that combines technical, business and customer facing app store properties. We applied our approach to 32,108 non-zero priced apps available from the Blackberry app store. Our results show that there is a strong correlation between customer rating and the rank of app downloads, though perhaps surprisingly, there is no correlation between price and downloads, nor between price and rating. We provide empirical evidence that our extracted features are meaningful and valuable: they maintain correlations observed at the app level and provide the input to price prediction system that we construct using Case Based Reasoning. Our prediction system statistically significantly outperforms recommended existing approaches to price estimation (and with at least medium effect size) in 16 out of 17 of Blackberry App Store categories.