Association rules in general show correlations among product items, e.g. the likelihood to buy battery and milk powder together. In practice, we look for causality, for example, if a customer buys Book A, what is the likelihood for this customer to buy Book B? If we use a transaction data, we don't know if it is correlation or a causality. Please give an actionable solution for a merchant to assess the causality for an event, such as those customers who have read Book A, they are likely interested in Book B and they will buy it. Note: We focus on transactional data. You are not expected to ask the merchant to send a survey or do some interviews with customers. Word limit: 100 words.f