| Title |
Criteria for Evaluating Characteristics of Data Models for Structural Experiment Information |
| Keywords |
Data Model ; Structural Experiment Information ; Lehigh Model ; NEES ; Class ; Object |
| Abstract |
The information involved in the structural experiments can be formally organized using the date models before the computer systems are implemented. A number of data models have been developed by the previous research efforts, such as the NEES data model for general information of structural experiments and the Lehigh Model for the large-scale and hybrid structural experiments. The comparisons and evaluations for the data models have been made only by the descriptive way, not by the numerical way. This paper proposes some numerical criteria for evaluating characteristics of data models for structural experiments. The criteria for the classes in a data model includes the total number of classes, the number of attributes in class, and the average number of attributes in class to describe the size and volume of the data model. The criteria for the objects includes the ratio of objects to classes, the ratio of attribute value existence in class, and the number of selections for value to describe the use of the data models for a specific project and to describe the complexity of the paths for specific information in the data model. These criteria are applied for the Lehigh model and the NEES data model, and the numerical values for the criteria are used for describing the characteristics of the two date models. |