NASA93
Author admin | 12.11.2007 | Category Dan Port, Effort Prediction, Jairus Hihn, Karen Lum, KortePort08, Marcel Korte, MenziesChenHihnLum06a, NASA, PROMISE 2008 - Discussion 1, Tim Menzies, Zhihao Chen
- Category:
- effort prediction
- Download:
- data/nasa93/nasa93.arff
- Donor:
- Tim Menzies
- Donated:
- February 8, 2006
- Source:
- Data from different centers for 93 NASA projects from 1980s and 1990s was collected by Jairus Hihn at JPL NASA Manager SQIP Measurement & Benchmarking Elements. Note: superseded by the nasa93 dataset
- Used by:
- [MenziesChenHihnLum06a ]
cocomonasa_v1
Author admin | 05.11.2007 | Category Dan Port, Effort Prediction, KortePort08, Marcel Korte, NASA, PROMISE 2008 - Discussion 1, Tim Menzies
- Category:
- effort prediction
- Download:
- data/nasa93/cocomonasa_v1.arff
- Donor:
- Tim Menzies
- Donated:
- December 2, 2004
- Source:
- NASA
About This Site
We seek repeatable, improvable, maybe even refutable, software engineering experiments.
To this end, we made this site where researchers can publish their data and the tools used to make their conclusions.
Caveat Emptor
An excessive focus on empirical results can stunt the development of innovative ideas that are, as yet, pre-experimental.
However, currently, the field of software engineering is in no danger of an excess of empiricism.
Repository
-
In 2006, the repository held 23 data sets.
- Defect Prediction (90)
- Effort Prediction (18)
- General (9)
- Model-based SE (8)
- Text Mining (9)
In 2008, at last update, the repository holds 134 data sets in the following areas:
Further contributions are always welcome in the above areas, or any other.
Why so much data? Firstly, there is the open source effect: public code and public logs means more data sets.
Secondly, the nature of an SE project means that once a tracking system is in place, then each new project (and each new release of each project) generates yet another data set.