生涯向來涉及人與環境兩類範疇，當人們出入於高中、大學與職業等不同世界而發生問題時，不僅在考驗著生涯領域的理論與工具，也在考驗著整體社會的生涯資訊基礎結構。台灣產官學各界都有不少相關調查、資料庫與常模，然而在分類系統等基礎結構不佳下，這些資料卻無法串接、更遑論整合。本研究意圖在修訂大考中心的學系學類分類之同時，以學類為核心，向上串接起高中輔導所用的十八學群、向下串接起教育部學科標準分類之系類、水平串接至美國學程分類系統 CIP之學程、以及跨界地串接起職業標準分類 O*NET-SOC之職業。在質性與量化兼具的兩階段研究設計下，修訂後學類與學職對應關係皆有令人滿意的效度水準，學類間關係樣態為環狀網絡結構，學類對應職業之中位數為 9種職業，除 530條 CIP職業對應外，亦增補了 673條學職對應關係。最後，本研究以學類關係描述子來統合與表徵研究結果，並討論了各關係描述子之可能應用。

Career involves people and their environment. Individuals exist and go through different systems, such as high school, university and the workplace. If career information is not well integrated through the different systems, individuals will have more difficulties in their career choices and transition. A classification system is the infrastructure of information connection and integration. Many career data and surveys are available in Taiwan. However, the information system lacks a basic framework, and therefore the data lacks interconnection and integration. This research sought to: (1)revise the classification of program fields in theTaiwan College Entrance Examination Center (CEEC), and based on program fields, up-connect it with 18 educational classifications used by high school guidance, down-connect it to standard field classifications delineated by the Ministry of Education, and cross-connect it to the United State’s Classification of Instructional Program (CIP); and (2)establish interconnection between programs delineated in the CEEC and O*NET-SOC, an occupational coding system. The research design comprised a two-stage quantitative and qualitative design. Discriminant analysis and the correlation analysis showed that the revised CEEC classification system and interconnection among between program groups and occupations achieved validity. The relational network among the program groups manifested a circular structure. The median number of occupations per program group was 9. In addition to the 530 data corresponding to the CIP occupation, 673 corresponding data between academic programs and occupation were added. Last, the research results were summarized and represented by using program relational descriptors, and applications for these descriptors were discussed.