Error propagation in distributed databases
CIKM '95 Proceedings of the fourth international conference on Information and knowledge management
A Hybrid Representation of Vague Collections for Distributed Object Management Systems
IEEE Transactions on Knowledge and Data Engineering
A Closed Approach to Vague Collections in Partly Inaccessible Distributed Databases
ADBIS '99 Proceedings of the Third East European Conference on Advances in Databases and Information Systems
Modelling uncertainty in multimedia database systems: an extended possibilistic approach
International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems
Incomplete information in multidimensional databases
Multidimensional databases
POP algorithm: Kernel-based imputation to treat missing values in knowledge discovery from databases
Expert Systems with Applications: An International Journal
Kernel-Based Multi-Imputation for Missing Data
Proceedings of the 2006 conference on Advances in Intelligent IT: Active Media Technology 2006
Optimized parameters for missing data imputation
PRICAI'06 Proceedings of the 9th Pacific Rim international conference on Artificial intelligence
Missing value imputation based on data clustering
Transactions on computational science I
Missing data imputation by utilizing information within incomplete instances
Journal of Systems and Software
| Hi-index | 0.00 |
This paper proposes a method for handling inapplicable and unknown missing data. The method is based on: (1) storing default values (instead of null values) in place of missing data, (2) storing truth values that describe the logical status of the default values in corresponding fields of corresponding tables. Four valued logic is used so that the logical status of the default data values can be described as, not just true or false, but also as inapplicable or unknown. This method, in contrast to the “hidden byte” approach, has two important advantages: (1) Because the logical status of all data is represented explicitly in tables, all 4-valued operations can be handled via a 2-valued data manipulation language, such as SQL. Language extensions for handling missing data (e.g., “IS NULL”) are not necessary. (2) Because data fields always contain a default value (as opposed to a null value or mark), it is possible to do arithmetic across missing data and to interpret the logical status of the result by means of logical operations on the corresponding stored truth values.