Research on design and implementation of corporate financial information database

在第一年，我们将“日经需求数据库”与“标准普尔研究洞察”在内容、用途和可用性方面进行了比较。日经需求数据库的界面相对简单，其搜索条件仅限于一些重要变量。Research Insight提供了多种方法，其搜索方法根据客户的目的进行分类。他们提供了一个“门户”，有四种搜索方式：(1)研究，(2)报告，(3)图表，(4)公司信息。同时，对实现的理论背景进行了研究。在实现阶段，我们必须提醒可用性。可用性包括界面和语义。我们采用XML技术和关系数据库管理系统来解决这两个问题。在用XML表示会计数据时，我们必须仔细分析会计项目。我们将Nikkei NEEDS财务文件数据分解为树状结构。第二年，我们将会计项目部署到XBRL（扩展业务报告语言）中，并将这些数据放入关系数据库中。我们还为此数据库实现了必要的接口（前端），并提供给早稻田大学的用户。这个数据库的性能非常快。通常的用法，例如搜索2000家公司的10个变量在10秒内结束。我们还使用该系统进行了几项实证研究。我们将根据用户的反馈继续改进系统。

In the first year, we compared "Nikkei NEEDS Database" with "Standards & Poor's Research Insight", in its contents, usage, and usability. Nikkei NEEDS database has a relatively simple interface, with its search criteria limited to some important variables. Research Insight offers a variety of methods, and its search method is categorized with customers' purposes. They offer a "portal", featuring four types of search methods: (1) research, (2) report, (3) graph, and (4) company information. In the meanwhile, we studied the theoretical background of implementation. In the implementation phase, we have to remind the usability. Usability consists of interface and semantics. We adopted XML technology and Relational Database Management System to solve these two challenges. In expressing accounting data with XML, we have to analyze account items carefully. We broke Nikkei NEEDS financial file data into tree structure.In the second year, we deployed accounting items into XBRL (Extensive Business Reporting Language) and put those data into the relational database. We also implemented necessary interface (front-end) for this database, and offered to the users at Waseda University.Performance of this database is quite fast. Usual usage, such as searching 10 variables for 2,000 companies ends within 10 seconds. We also used this system for several empirical studies. We are continuing to work to improve the system with the feed back from users.