Database design for error searching system based on keyword priority
Fan Yang, Zheng-hong Dong, Zhi‐Wei Liu · 2016
Because there are too many types of errors occurred in a multi-software integrated platform, such as Integrated Decision Information System (IDIS). It is urgent to design an error searching system to solve all different problems. However, those errors belong to different stages like setup, configuration, and operation, or those errors may occurred in different services, applications, or IP ports, or may be happened in different system software, different version of software, and those errors are also can be classified into different types. The new requirement on the design of database has been announced, that it has to locate the error, find out the reason of this error, as well as the corresponding solution. Also, it requires to provide the location, phase, etc. of an error. For a multi-software integrated platform, this paper proposed a database design for error searching system based on keyword priority. This DB made the correspondence among error, the reason of the error, and corresponding solution, and put them to different categories in terms of their characteristics, such that it is easy to manage, search, and use. This method treats those characteristics as keywords with higher priority, which are correspondent with errors, reasons, and solutions, respectively. While the keywords extracted from errors, reasons, and solutions are treated as keywords with lower priority. Keywords with different priorities can all be used as index to search errors, reasons, and solutions, while keywords with higher priority can be used to filter the searching results finer, and make the searching results more accurate. On one hand, the database design method based on keyword priority has simplified the logical structure of the database. On the other hand, users do not only search the reason and solution of an error, but also can find out the accurate information of the error, such as in which stage, layer, software, or categories, etc. the error has happened. The data of the database is complete, which has been provided by 500 technicians who had found thousands of errors. Those errors haven been added to the DB after tests to make the DB more complete.