對於不同團隊協同開發大型的應用程式，常會面臨應用程式自行處理例外事件而未正確回報用戶之執行結果與無法得知封閉式類別庫內部詳細呼叫流程，導致偵錯不易的困擾。有鑒於此，本研究將實作一個工具，利用Aspect-Oriented Programming(AOP)特性，將AspectJ織入JDBC驅動程式並賦予存取紀錄儲存之功能，既可免除應用程式本身未針對資料庫儲存存取紀錄的問題，直接儲存所有存取紀錄，亦不影響資料庫操作，更可以分辦造成錯誤發生的程式以釐清問題，有效幫助偵錯。此工具適用於任何使用JDBC介面去存取資料庫管理系統（DBMS）之JAVA應用程式，它能在不影響應用程式的情況下客製化儲存應用程式透過JDBC驅動程式連結資料庫管理系統（DBMS）的存取紀錄，並利用存取紀錄進行分析並複製存取行為，以提供更豐富的資訊進一步協助偵錯。 / When using a large database application which is jointly designed by different teams, the administrators are often confronted with a problem that the application handles the exception spontaneously and often fails to report back the result of user’s execution. Therefore, the administrator can’t be informed of the details of the calling flow in case of exceptions. This will make it not only difficult to find which part of the application goes wrong, but also hard to fix. To resolve this problem, this thesis designs and implements an aspect-based debugging tool for such database applications. The key idea behind the tool is that, instead of adapting the large application for logging database access activities, we target at the JDBC driver and make JDBC driver capable of saving access log. This tool works for any JAVA applications that use JDBC interface to access database. Without any modification to the user-level application, this tool can save the access log obtained by application which use JDBC driver to connect the database, and the administrator can customize the access log. By analyzing the access records and duplicating the access behavior, this tool provides more information to further assist in application debugging.

領域專屬語言的設計理念是希望能夠協助使用者在特定領域上解決特定問題。然而，大多數領域專屬語言的開發環境與工具均非常貧乏，這將會增加使用者在開發程式上的困難。 所以本研究利用模型導向技術來建構一套生成系統，使用者只要輸入領域專屬語言的語意以及偵錯定義，就能快速生成語言的執行與偵錯環境，並提供完善的操作介面，輔助使用者加快程式開發的速度。 / The purpose of creating domain-specific languages (DSLs) is to help user solve problems in a particular domain. However, most DSLs are lack of development environments and tools, which would be more difficult to develop programs. This thesis is aimed to develop a generating system based on model-driven technology. Given the semantics and debugging definitions of a domain-specific language, the system would be able to generate an executing and debugging environment for the language with friendly user interface, thus improving efficiency and productivity of using the language.

現今變化快速的環境下，科技團隊面對研發過程中日益增高的失敗率，失敗帶來打擊但也可能是創新的來源。然而，團隊創新的研究卻少有團隊失敗、錯誤與創新關係的研究。有鑑於此，本研究以國內485位研發工程師（某法人單位35個團隊共323位成員、科技產業38個團隊共162位成員）共計73個團隊為研究對象。本研究檢視錯誤管理實務（團隊偵錯能力與團隊錯誤溝通能力）與團隊創新績效之間的關係、錯誤管理實務與錯誤管理氛圍（錯誤學習信念與情緒）的關係、錯誤管理氛圍與前導因子（鼓勵實驗、教導型領導、目標清晰度、衝突處理方式）的關係。結果發現：(1）團隊偵錯能力越高時，團隊創新績效越高；(2）群體從錯誤中學習信念越高時，團隊偵錯能力越高；(3）團隊鼓勵小型實驗與主管進行教導型領導時，群體錯誤學習信念越高。顯示錯誤管理是團隊創新中不可忽視的一環。 　　另本研究發現教導型領導、鼓勵實驗對團隊形成錯誤學習信念有正向影響，與社會認知理論呼應。本研究也發現，團隊之年資多樣性與錯誤處理情緒與錯誤溝通能力皆呈負向影響。 / Extending previous research on team error management, this thesis is conducted to examine the antecedents, error management climate, error management practice and consequences of team innovation. Data is collected from 35 R&D teams (an anonymous government research institutions in Taiwan) and 38 R&D teams (Top 1000 Technology enterprises in Taiwan). We give the following three hypotheses. First, we hypothesize that detecting capability and misunderstanding communication skills influence team innovation performance. Second, we hypothesize that error management climate (shared belief and emotion) influence detecting capability and misunderstanding communication skills. Third, we hypothesize that effective coaching, clear direction, conflict management and encouragement of experiment influence error management culture (belief and emotion). The results of structural equation model analysis revealed that detecting capability positively predicts team innovation performance. Shared belief positively predicts team detecting capability. The results also show effective coaching and encouragement of experiment positively predict shared belief which is extended from social cognitive theory. Results of the analysis also indicate that tenure diversity negatively predicts error management emotion and misunderstanding communication skills.