Paper Details

The current era is witnessing a vast development in all fields of animal health care. The usage of IT to transfer information and knowledge in the animal health care domain using expert systems is one of the areas investigated by many institutions. The transfer of knowledge from veterinary consultants & scientists to livestock holders represents a bottleneck for the development of animal health care in any country. Measuring the performance of an expert system is difficult as knowledge cannot be quantified. But it is impossible to present formal proofs of correctness for expert systems. Expert systems can be supported with mechanism for correctness of the results for users who rely on them. The paper has proposed a new kind of association algorithm based on support and certainty, which can scan the datasets several times to discover certain frequent item sets whose length complied with fixed increment. The objective of this paper is to propose a data mining technique, which can be used for storage & assessment of data for different diseases in lactating animals and observe associative rules based on clinical diagnostic parameters. The dataset being designed in this study for storage of clinical data, should be able to render the data in appropriate structures, provide metadata that adequately records semantics of data. Applying Apriori algorithm to a given clinical data set can have the potential to confirm results of expert system regarding disease and symptoms co-occurrences. Under the architecture based on support and confidence, the Apriori algorithm can prove to be certainty association rule, when a piece of association rule is larger than the expected support and meets the condition of certainty.

Data mining, Apriori Algorithm, Expert System, Lactating Animals