2015 13th International Conference on Frontiers of Information Technology (FIT) (2015)
Dec. 14, 2015 to Dec. 16, 2015
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/FIT.2015.32
Database integration is a process based on user requirement for unified view of data from heterogeneous data sources. Conflicts are necessary to resolve for successful integration. Database integration may result in the data level conflicts and some of the conflicts cause the inappropriate representation of data. These conflicts mostly arise after the schema integration process which is performed prior to the integration of data. One of the very common conflicts is null value conflict mostly caused by missing attributes value in global table of integrated schema. Most of the schema integration techniques add the dummy attributes in global table for missing attribute in local database which may result in this conflict. In this paper we discuss different cases of null value conflicts and its occurrence after schema integration. We also represent this conflict formally using Z-notation for the first time. We provide the formal resolution function for this conflict as well. Algorithm of this resolution function is also implemented and we developed a system called DHResol, this system resolves the null value conflict as well as integrates the data form different input databases.
Null value, Databases, Data integration, Merging, Semantics, Uncertainty
G. A. Mirza, "Null Value Conflict: Formal Definition and Resolution," 2015 13th International Conference on Frontiers of Information Technology (FIT), Islamabad, Pakistan, 2015, pp. 132-137.