METHOD AND APPARATUS FOR CALCULATING RECOMMENDATION SCORE BASED ON CONDITIONAL PROBABILITY

摘要

The present invention relates to a conditional probability-based product recommendation score calculation method and apparatus, comprising: a plurality of first products sold in a first preset period and a plurality of second products sold in a second preset period after the first period A binary data value corresponding to a product-specific purchase by a first customer who has purchased at least one of the first product and the second product among products including, and whether a second customer who has purchased at least one of the first product has a product-by-product purchase. Storing binary data values corresponding to each of, and binary data values corresponding to whether the second customer has purchased the second product with respect to binary data values corresponding to whether the first customer has purchased the first product. And generating a conditional probability based on the number of items, and calculating a recommendation score for the second product of the second customer based on the conditional probability. Accordingly, the recommendation score is calculated based on the prediction of whether the second customer has purchased the second product based on the conditional probability generated based on whether the first customer has purchased the first product and whether the second product has been purchased with respect to the second customer. Thus, there is an effect that can improve the recommended performance.