Skip to main contentCal State San Bernardino
>> [CNS] >> [Comp Sci Dept] >> [R J Botting] >> [New Bibliographic Items] >> newb0124 [Blog/News] || [Purpose] || [Notation] || [Copyright] || [Site Search] || [Bibliography Search]
Wed Jan 25 16:32:27 PST 2006

Contents


    SawyerRaysonCosh05

    1. Pete Sawyer & Paul Rayson & Ken Cosh
    2. Shallow knowledge as an aid to deep understanding in early phase requirements Engineering
    3. IEEE Trans Software Engineering V31n11(Nov 2005)pp969-981
    4. =DEMO STATISTICAL TEXT ANALYSIS REQUIREMENTS ATC GLOSSARY ONTOLOGY SYSTEMS DOMAIN MODEL
    5. Simple stats leads to useful conclusions when analyzing current systems.
    6. (dick) |- Figure 3 commits several UML sins.

    Jorgensen05

    1. Magne Jorgensen
    2. Evidence-Based Guidelines for assessment of Software Development Cost Uncertainty
    3. IEEE Trans Software Engineering V31n11(Nov 2005)pp942-954
    4. =SURVEY DEVELOPMENT COST ESTIMATION
    5. Uncovers evidence that software developers use intuition and expertise to predict costs and tend to underestimate them as a result.
    6. Propose some guidelines based on on a very thorough literature survey.
    7. Examples:
      • Don't ask for 90% confidence intervals. Instead ask fro an interval and then ask what % of projects were outside the interval.
      • Formal cost estimation are currently less efficient - intervals are unusable.
      • Use a structured process with explicit reasons.
      • Compare this project with previous projects to estimate errors.
      • Perhaps use groups rather than formulas to combine estimates.
      • Paying people if they are more accurate may or may not improve their accuracy.

    LiuFeketeGorton05

    1. Yan Liu & Alan Fekete & Ian Gorton
    2. Design-level Performance Prediction of Component-Base Applications
    3. IEEE Trans Software Engineering V31n11(Nov 2005)pp928-941
    4. =SIMULATION =THEORY QUEUE COMPONENTS PERFORMANCE EJB QNM
    5. Derives a queuing theory performance model for various kinds of transactions.
    6. Models calibrated by a very simple benchmark on two platforms.
    7. Models combined with a sequence diagram model of an application to predict performance.
    8. The predictions compares well with a simulation of the application.
    9. The predictions allow the selection of the better of two possible designs.
    10. No code is needed in to make the predictions.
    11. (dick) |- SSADM Physical Design Control used quantified models of DBMSs and application designs to select a physical design that was likely to perform well in 1979.

    Lin06

    1. Frank Lin
    2. The concerns of innovation in organizations: a comparison of managerial and end-user perspectives and an Individual's Stages of concern
    3. IDS Seminar, CBPA CSUSB (Jan 20 2006)
    4. =EXPERIENCE CHINA PHARMACEUTICAL ERP
    5. Senior manager see things differently.
    6. China less positive than US.
    7. Concern-Based Adoption Model,
    8. CBAM::= awareness; informational; personal; management; consequence; collaboration; refocusing.

    Wilson06

    1. Gregory V Wilson
    2. Where's the Real Bottleneck in Scientific Computing
    3. American Scientist V94n1(Jan-Feb 2006)pp5-6
    4. =EXPERIENCE TOOLS COMPUTATIONAL SCIENCE IDE UNIT TESTS CM EDUCATION
    5. "the bottleneck between our ears"
    6. Claims scientists need training in using integrated development environments, symbolic debuggers, unit testing, reading code, version control, automating repetitive tasks, scripting languages, ...
    7. On the web [ swc ]

End