The Prediction Model of Case-Based Reasoning about Project Evaluation of Xiamen City

This paper was based on construction cost indexes of Xiamen, 11 feature attributes were extracted, including engineering purposes, engineering categories, and 3 predictors including square metre cost. Frame-based representation was used to represent cases. Access was used to build engineering case library. Visual Basic language was adopted to develop object-oriented design. Combined with knowledge guidance strategy and the nearest neighbor algorithm, hybrid retrieval strategy was adopted in basic retrieval and advanced retrieval. When built on case-based reasoning, engineering evaluation model provids a fast and effective method for determining the investment target of project planning and feasibility studies.