The majority of information retrieval experiments are evaluated by measures such as average precision and average recall. Fundamental decisions about the superiority of one retrieval technique over another are made solely on the basis of these measures. We claim that average performance figures need to be validated with a careful statistical analysis and that there is a great deal of additional information that can be uncovered by looking closely at the results of individual queries. This paper is a case study of stemming algorithms which describes a number of novel approaches to evaluation and demonstrates their value.

Software requirements specifications (SRS) are often validated manually. One such process is inspection, in which several reviewers independently analyze all or part of the specification and search for faults. These faults are then collected at a meeting of the reviewers and author(s). Usually, reviewers use Ad Hoc or Checklist methods to uncover faults. These methods force all reviewers to rely on nonsystematic techniques to search for a wide variety of faults. We hypothesize that a Scenario-based method, in which each reviewer uses different, systematic techniques to search for different, specific classes of faults, will have a significantly higher success rate. We evaluated this hypothesis using a 3 \Theta 2 4 partial factorial, randomized experimental design. Forty eight graduate students in computer science participated in the experiment. They were assembled into sixteen, three-person teams. Each team inspected two SRS using some combination of Ad Hoc, Checklist or Scenario meth...

Model selection is important in many areas of supervised learning. Given a dataset and a set of models for predicting with that dataset, we must choose the model which is expected to best predict future data. In some situations, such as online learning for control of robots or factories, data is cheap and human expertise costly. Cross validation can then be a highly effective method for automatic model selection. Large scale cross validation search can, however, be computationally expensive. This paper introduces new algorithms to reduce the computational burden of such searches. We show how experimental design methods can achieve this, using a technique similar to a Bayesian version of Kaelbling’s Interval Estimation. Several improvements are then given, including (1) the use of blocking to quickly spot near-identical models, and (2) schemata search: a new method for quickly finding families of relevant features. Experiments are presented for robot data and noisy synthetic datasets. The new algorithms speed up computation without sacrificing reliability, and in some cases are more reliable than conventional techniques. 1

Traffic anomalies such as failures and attacks are commonplace in today's network, and identifying them rapidly and accurately is critical for large network operators. The detection typically treats the traffic as a collection of flows that need to be examined for significant changes in traffic pattern (e.g., volume, number of connections) . However, as link speeds and the number of flows increase, keeping per-flow state is either too expensive or too slow. We propose building compact summaries of the traffic data using the notion of sketches. We have designed a variant of the sketch data structure, k-ary sketch, which uses a constant, small amount of memory, and has constant per-record update and reconstruction cost. Its linearity property enables us to summarize traffic at various levels. We then implement a variety of time series forecast models (ARIMA, Holt-Winters, etc.) on top of such summaries and detect significant changes by looking for flows with large forecast errors. We also present heuristics for automatically configuring the model parameters. Using a

propose a preliminary set of research guidelines aimed at stimulating discussion among software researchers. They are based on a review of research guidelines developed for medical researchers and on our own experience in doing and reviewing software engineering research. The guidelines are intended to assist researchers, reviewers, and meta-analysts in designing, conducting, and evaluating empirical studies. Editorial boards of software engineering journals may wish to use our recommendations as a basis for developing guidelines for reviewers and for framing policies for dealing with the design, data collection, and analysis and reporting of empirical studies. Index TermsÐEmpirical software research, research guidelines, statistical mistakes. 1

Solutions to the problem of learning from examples will have far-reaching benefits, and therefore, the problem is one of the most widely studied in the field of machine learning. The purpose of this study is to investigate a general solution method for the problem, the automatic generation of decision (or classification) trees. The approach is to provide insights through in-depth empirical characterization and evaluation of decision trees for one problem domain, software resource data analysis. The purpose of the decision trees is to identify classes of objects (software modules) that had high development effort or faults, where "high" was defined to be in the uppermost quartile relative to past data. Sixteen software systems ranging from 3000 to 112,000 source lines have been selected for analysis from a NASA production environment. The collection and analysis of 74 attributes (or metrics), for over 4700 objects, capture a multitude of information about the objects: development effort...

Software review is a fundamental tool for software quality assurance. Nevertheless, there are significant controversies as to the most efficient and effective review method. One of the most important questions currently being debated is the utility of meetings. Although almost all industrial review methods are centered around the inspection meeting, recent findings call their value into question. To gain insight into these issues, the two authors of this paper separately and independently conducted controlled experimental studies. This paper discusses a joint effort to understand the broader implications of these two studies. To do this, we designed and carried out a process of "reconciliation" in which we established a common framework for the comparison of the two experimental studies, re-analyzed the experimental data with respect to this common framework, and compared the results. Through this process we found many striking similarities between the results of the two studies, stren...

Previous proposals for soft-error tolerance have called for redundantly executing a program as two concurrent threads on a superscalar microarchitecture. In a balanced superscalar design, the extra workload from redundant execution induces a severe performance penalty due to increased contention for resources throughout the datapath. This paper identifies and analyzes four key factors that affect the performance of redundant execution, namely 1) issue bandwidth and functional unit contention, 2) issue queue and reorder buffer capacity contention, 3) decode and retirement bandwidth contention, and 4) coupling between redundant threads' dynamic resource requirements. Based on this analysis, we propose the SHREC microarchitecture for asymmetric and staggered redundant execution. This microarchitecture addresses the four factors in an integrated design without requiring prohibitive additional hardware resources. In comparison to conventional single-threaded execution on a state-ofthe -art superscalar microarchitecture with comparable cost, SHREC reduces the average performance penalty to within 4% on integer and 15% on floating-point SPEC2K benchmarks by sharing resources more efficiently between the redundant threads.

Given a set of models and some training data, we would like to find the model that best describes the data. Finding the model with the lowest generalization error is a computationally expensive process, especially if the number of testing points is high or if the number of models is large. Optimization techniques such as hill climbing or genetic algorithms are helpful but can end up with a model that is arbitrarily worse than the best one or cannot be used because there is no distance metric on the space of discrete models. In this paper we develop a technique called "racing" that tests the set of models in parallel, quickly discards those models that are clearly inferior and concentrates the computational effort on differentiating among the better models. Racing is especially suitable for selecting among lazy learners since training requires negligible expense, and incremental testing using leave-one-out cross validation is efficient. We use racing to select among various lazy learnin...

We replicated a controlled experiment first run in the early 1980's to evaluate the effectiveness and efficiency of 50 student subjects who used three defect-detection techniques to observe failures and isolate faults in small C programs. The three techniques were code reading by stepwise abstraction, functional (black-box) testing, and structural (white-box) testing. Two internal replications showed that our relatively inexperienced subjects were similarly effective at observing failures and isolating faults with all three techniques. However, our subjects were most efficient at both tasks when they used functional testing. Some significant differences among the techniques in their effectiveness at isolating faults of different types were seen. These results suggest that inexperienced subjects can apply a formal verification technique (code reading) as effectively as an execution-based validation technique, but they are most efficient when using functional testing.