One of the most common types of cancer that affects women is breast cancer. The early detection is very important for this disease. Eventhough genetic mutations and family history are associated with the high risk of developing breast cancer there can be other factors also which decide the chances for the breast cancer in women. For eg. aggreviated levels of CEA(Carcino embrionic Antigen) can be a pre-indication to the cancer and is associated with metastatic disease. However with a single biomarker or a medical attribute, the diagnosis may not be done appropriately and a combination of biomarkers or attributes related to the disease can certainly contribute in the prediction of breast cancer in an early stage. The selection of the appropriate combination of these medical attributes is a challenge as the knowledge about the best combination of these attributes will help the medical field experts and doctors to focus more on these findings and determine treatments based on these findings. Various association rule mining techniques like classic Apriori Algorithm are endorsed to explore the best combinations of these attributes. But one of the major disadvantages of these traditional algorithms are,dataset need to be passed more than two times or atleast two times to identify the repeating and significant attributes. Therefore a cogency based association rule mining algorithm called CbARM algorithm is proposed and applied to the SEER dataset 2020(Nov). The concept of cogency refers to the likelihood that the presumed facts will be true if the conclusion is correct. One of the advantages of the proposed method is that it can be used to generate rules with just a single pass of the file. This is made possible by the construction of a knowledge link matrix. It is found that compared to Apriori algorithm more suitable combinations of medical attributes in terms of rules have been generated , once the CbARM algorithm is applied.

Breast cancer, Survivability, Association Rule Mining