Precision prediction based on ranked list coherence   总被引：1，自引：0，他引：1  

Steve Cronen-Townsend  Yun Zhou  W. Bruce Croft 《Information Retrieval》2006,9(6):723-755

We introduce a statistical measure of the coherence of a list of documents called the clarity score. Starting with a document list ranked by the query-likelihood retrieval model, we demonstrate the score's relationship to query ambiguity with respect to the collection. We also show that the clarity score is correlated with the average precision of a query and lay the groundwork for useful predictions by discussing a method of setting decision thresholds automatically. We then show that passage-based clarity scores correlate with average-precision measures of ranked lists of passages, where a passage is judged relevant if it contains correct answer text, which extends the basic method to passage-based systems. Next, we introduce variants of document-based clarity scores to improve the robustness, applicability, and predictive ability of clarity scores. In particular, we introduce the ranked list clarity score that can be computed with only a ranked list of documents, and the weighted clarity score where query terms contribute more than other terms. Finally, we show an approach to predicting queries that perform poorly on query expansion that uses techniques expanding on the ideas presented earlier.

W. Bruce CroftEmail:

  相似文献   

Bayes not Bust! Why Simplicity is no Problem for Bayesians     

Dowe  David L.; Gardner  Steve; Oppy  Graham 《The British journal for the philosophy of science》2007,58(4):709-754

The advent of formal definitions of the simplicity of a theoryhas important implications for model selection. But what isthe best way to define simplicity? Forster and Sober ([1994])advocate the use of Akaike's Information Criterion (AIC), anon-Bayesian formalisation of the notion of simplicity. Thisforms an important part of their wider attack on Bayesianismin the philosophy of science. We defend a Bayesian alternative:the simplicity of a theory is to be characterised in terms ofWallace's Minimum Message Length (MML). We show that AIC isinadequate for many statistical problems where MML performswell. Whereas MML is always defined, AIC can be undefined. WhereasMML is not known ever to be statistically inconsistent, AICcan be. Even when defined and consistent, AIC performs worsethan MML on small sample sizes. MML is statistically invariantunder 1-to-1 re-parametrisation, thus avoiding a common criticismof Bayesian approaches. We also show that MML provides answersto many of Forster's objections to Bayesianism. Hence an importantpart of the attack on Bayesianism fails.

1. Introduction
2. TheCurve Fitting Problem

   2.1 Curves and families of curves

   2.2 Noise

   2.3 Themethod of Maximum Likelihood

   2.4 ML and over-fitting

3. Akaike's Information Criterion(AIC)
4. The Predictive Accuracy Framework
5. The Minimum MessageLength (MML) Principle

   5.1 The Strict MML estimator

   5.2 Anexample: Thebinomial distribution

   5.3 Properties ofthe SMML estimator

   5.3.1  Bayesianism

   5.3.2  Languageinvariance

   5.3.3Generality

   5.3.4  Consistencyand efficiency

   5.4 Similarity to false oracles

   5.5 Approximationsto SMML

6. Criticisms of AIC

   6.1 Problems with ML

   6.1.1  Smallsample biasin a Gaussian distribution

   6.1.2  Thevon Misescircular and von Mises—Fisherspherical distributions

   6.1.3  The Neyman–Scottproblem

   6.1.4  Neyman–Scott,predictive accuracyandminimum expected KL distance

   6.2 Otherproblems with AIC

   6.2.1  Univariate polynomial regression

   6.2.2  Autoregressiveeconometric time series

   6.2.3  Multivariatesecond-orderpolynomial modelselection

   6.2.4  Gapor no gap:a clustering-like problem forAIC

   6.3 Conclusionsfrom the comparison of MML and AIC

7. Meeting Forster's objectionsto Bayesianism

   7.1 The sub-family problem

   7.2 Theproblem of approximation,or, which framework forstatistics?

8. Conclusion

1. Details of the derivation of the Strict MMLestimator
2. MML, AIC and the Gap vs. No Gap Problem

   B.1 Expectedsize of the largest gap

   B.2 Performanceof AIC on thegap vs. no gap problem

   B.3 Performanceof MML in thegap vs. no gap problem

  相似文献   

Breadth and Balance? The Impact of the National Literacy and Numeracy Strategies on Art in the Primary School     

Steve Herne 《The International Journal of Art & Design Education》2000,19(2):217-223

This paper presents the findings of a small-scale qualitative research enquiry into some of the effects on the primary school art curriculum of the introduction of the National Literacy and Numeracy Strategies in 1988 and 1999. Five art curriculum co-ordinators and an additional part time specialist art teacher, drawn from five inner city primary schools in two London boroughs, were the subject of a semi-structured interview designed to elicit data on the broad changes in primary art education. Key external factors were the introduction of the Art National Curriculum in 1992, its subsequent development and its current condition during a period of ‘relaxation’ of the Orders, following the government's introduction of the Literacy and Numeracy Strategies. Little official attention has been given to other, perhaps unintended, outcomes of the latter strategy across the curriculum. This study gives an insight into the experience and perceptions of some of those carrying the responsibility for leadership of art in school in the primary phase.  相似文献   

Clarity, charity and criticism, wit, wisdom and worldliness: Avoiding intellectual impositions     

David Turnbull  Henry Krips  Val Dusek  Steve Fuller  Alan Sokal  Jean Bricmont  Alan Frost  Alan Chalmers  Anna Salleh  Alfred I. Tauber  Yvonne Luxford  Nicolaas Rupke  Steven French  Peter G. Brown  Hugh LaFollette  Peter Machamer  Nicolas Rasmussen  Andy J. Miller  Marya Schechtman  Ross S. West  John Forge  David Oldroyd  Nancy Demand  Darrin W. Belousek  Warren Schmaus  Sungook Hong  Alan Chalmers  Rachel A. Ankeny  Peter Anstey  Jeremy Butterfield  Harshi Gunawardena 《Metascience》2000,9(3):347-498

  相似文献   

Sequences of upper and lower extremity motions in javelin throwing     

Liu H  Leigh S  Yu B 《Journal of sports sciences》2010,28(13):1459-1467

Javelin throwing is technically demanding. Sequences of upper and lower extremity motions are important for javelin throwing performance. The purpose of this study was to determine the general sequences of upper and lower extremity motions of elite male and female javelin throwers. Three-dimensional kinematic data were collected for 32 female and 30 male elite javelin throwers during competitions. Shoulder, elbow, wrist, hip, knee, ankle, lower trunk, and upper trunk joint and segment angles were reduced for the best trial of each participant. Beginning times of 6 upper extremity and 10 lower extremity joint and segment angular motions were identified. Sequences of the upper and lower extremity motions were determined through statistical analyses. Upper and lower extremity motions of the male and female elite javelin throwers followed specific sequences (P?≤?0.050). Upper extremity motions of the male and female elite javelin throwers did not follow a proximal-to-distal sequence as suggested in the literature. Male and female elite javelin throwers apparently employed different sequences for upper and lower extremity motions (P?相似文献   

Reviews     

Kim Sterelny  John Forge  Ivan Crozier  Sverre Myhra  Randall Albury  Steve Clarke  Yvonne Luxford  David Philip Miller  Lynn K. Nyhart  Mary Chan  Richard McDonough  Peter J. Riggs  Allan Franklin  Robert Nola  David Bloor  Mark Cortiula  Thomas Nickles  David Oldroyd  Nicolas Rasmussen  John Forge  William A. S. Sarjeant 《Metascience》1998,7(2):331-418

  相似文献   

The Impact of School Disadvantage on Senior Secondary Science: a Study of Patterns of Participation and Achievement in Government Secondary Schools in Victoria,Australia     

Murphy  Steve 《Research in Science Education》2020,50(4):1603-1618

Research in Science Education - Science and science education are recognised internationally as essential for ensuring a sustainable and prosperous future. At the same time, significant equity...  相似文献   

An assessment of social sciences coverage by four prominent full-text online aggregated journal packages   总被引：1，自引：0，他引：1  

Steve Black   《Library Collections, Acquisitions, and Technical Services》1999,23(4):39

The cost effectiveness and quality of full-text journals are analyzed for four prominent online aggregated journal packages: EBSCOhost Academic Search FullTEXT, UMI Proquest Direct Periodicals Research II, IAC’s Expanded Academic ASAP, and H.W. Wilson’s OmniFile. Price data from EBSCO’s Librarians’ Handbook are used to assess the total and average value of social sciences journals in each package. Quality of social sciences journals coverage is compared based on citation impact factors as recorded in Journal Citation Reports—Social Sciences Edition.  相似文献   

The relationship between discus throwing performance and combinations of selected technical parameters     

Leigh S  Gross MT  Li L  Yu B 《Sports biomechanics / International Society of Biomechanics in Sports》2008,7(2):173-193

The aim of this study was to determine the relationship between discus throwing performance and the following technical parameters: hip-shoulder and shoulder-arm separation, trunk forward-backward tilt, throwing-arm elevation angles, and the absolute and relative throwing procedure phase times. Videographic data of 51 male and 53 female discus throwers' competitive performances were captured during major meets. The three-dimensional coordinates of 21 body landmarks and the discus were obtained for each thrower's best trial using direct linear transformation. The technical parameters were reduced from the three-dimensional data at six critical instants during the throwing procedure. Hierarchical stepwise multiple regression analyses were performed to determine the relative influence of linear combinations of the technical parameters on discus throwing performance. Specific techniques associated with linear combinations of certain technical parameters were identified using canonical correlations. Males and females were analysed separately. Suggestions for controlling the height of release and the vertical component of the speed of release using an effective technique are made.  相似文献   

Indoor 16.1-km time-trial performance in cyclists aged 25- 63 years     

Balmer J  Bird S  Davison R 《Journal of sports sciences》2008,26(1):57-62

In this study, we assessed age-related changes in indoor 16.1-km cycling time-trial performance in 40 competitive male cyclists aged 25-63 years. Participants completed two tests: (1) a maximal ramped Kingcycle ergometer test, with maximal ramped minute power (RMPmax, W) recorded as the highest mean external power during any 60 s and maximal heart rate (HRmax, beats min(-1)) as the highest value during the test; and (2) an indoor Kingcycle 16.1-km time-trial with mean external power output (W), heart rate (beats min(-1)), and pedal cadence (rev min(-1)) recorded throughout the event. Results revealed age-related declines (P < 0.05) in absolute and relative time-trial external power output [(24 W (7.0%) per decade], heart rate [7 beats min(-1) (3.87%) per decade], and cadence [3 rev min(-1) (3.1%) per decade]. No relationships (P > 0.05) were observed for mean power output and heart rate recorded during the time-trial versus age when expressed relative to maximal ramped minute power and maximal heart rate respectively. Strong relationships (P < 0.05) were observed for maximal ramped minute power and time-trial power (r= 0.95) and for maximal heart rate and time-trial heart rate (r= 0.95). Our results show that indoor 16.1-km time-trial performance declines with age but relative exercise intensity (%RMPmax and %HRmax) does not change.  相似文献