Existing research is unclear on how to generate lessons learned for defect prediction and effort estimation. Should we seek lessons that are global to multiple projects or just local to particular projects? This paper aims to comparatively evaluate local versus global lessons learned for effort estimation and defect prediction. We applied automated clustering tools to effort and defect datasets from the PROMISE repository. Rule learners generated lessons learned from all the data, from local projects, or just from each cluster.The results indicate that the lessons learned after combining small parts of different data sources (i.e., the clusters) were superior to either generalizations formed over all the data or local lessons formed from particular projects. We conclude that when researchers attempt to draw lessons from some historical data source, they should 1) ignore any existing local divisions into multiple sources, 2) cluster across all available data, then 3) restrict the learning of lessons to the clusters from other sources that are nearest to the test data.

Abstract-Data miners can infer rules showing how to improve either (a) the effort estimates of a project or (b) the defect predictions of a software module. Such studies often exhibit conclusion instability regarding what is the most effective action for different projects or modules.This instability can be explained by data heterogeneity. We show that effort and defect data contain many local regions with markedly different properties to the global space. In other words, what appears to be useful in a global context is often irrelevant for particular local contexts.This result raises questions about the generality of conclusions from empirical SE. At the very least, SE researchers should test if their supposedly general conclusions are valid within subsets of their data. At the very most, empirical SE should become a search for local regions with similar properties (and conclusions should be constrained to just those regions).

Diversity in nomenclature and on-going dilemmas over the conceptual bases for the classification of voice disorders make it virtually impossible for the collation and accurate comparison of evidence-based data across different clinical settings. This has significant implications for treatment outcome studies. The first aim of this study was to develop a modified diagnostic classification system for voice disorders with clearly defined operational guidelines by which we might reliably distinguish voice disorders from one another. The second aim was to establish the face validity and reliability of the system as an effective diagnostic tool for the allocation of patients to different diagnostic groups for clinical and research purposes. After the Diagnostic Classification System for Voice Disorders (DCSVD) had been developed, it was used in an inter-rater reliability study for the independent assessment of 53 new consecutive patients referred to the Voice Analysis Clinics of three tertiary hospitals. There were three raters present for the assessment and diagnostic allocation of each patient. The high levels of inter-rater reliability suggest this may be a robust classification system that has good face validity and even at this early stage, strong construct validity.

No abstract

Context. M 87 is a giant elliptical galaxy located in the centre of the Virgo cluster, which harbours a supermassive black hole of mass 6.4 × 10 9 M , whose activity is responsible for the extended (80 kpc) radio lobes that surround the galaxy. The energy generated by matter falling onto the central black hole is ejected and transferred to the intra-cluster medium via a relativistic jet and morphologically complex systems of buoyant bubbles, which rise towards the edges of the extended halo. Aims. To place constraints on past activity cycles of the active nucleus, images of M 87 were produced at low radio frequencies never explored before at these high spatial resolution and dynamic range. To disentangle different synchrotron models and place constraints on source magnetic field, age and energetics, we also performed a detailed spectral analysis of M 87 extended radio-halo. Methods. We present the first observations made with the new Low-Frequency Array (LOFAR) of M 87 at frequencies down to 20 MHz. Three observations were conducted, at 15−30 MHz, 30−77 MHz and 116−162 MHz. We used these observations together with archival data to produce a low-frequency spectral index map and to perform a spectral analysis in the wide frequency range 30 MHz-10 GHz. Results. We do not find any sign of new extended emissions; on the contrary the source appears well confined by the high pressure of the intracluster medium. A continuous injection of relativistic electrons is the model that best fits our data, and provides a scenario in which the lobes are still supplied by fresh relativistic particles from the active galactic nuclei. We suggest that the discrepancy between the low-frequency radiospectral slope in the core and in the halo implies a strong adiabatic expansion of the plasma as soon as it leaves the core area. The extended halo has an equipartition magnetic field strength of 10 μG, which increases to 13 μG in the zones where the particle flows are more active. The continuous injection model for synchrotron ageing provides an age for the halo of 40 Myr, which in turn provides a jet kinetic power of 6−10 × 10 44 erg s −1 .