Step 1: Text relevance; text relevance refers to whether the text description information of the product (including: product title, category name, brand name, book products, authors, publishers, etc.) and the search keywords are relevant or matched.

Step 2: Category; whether the category (category of the product) to which the product belongs is reasonable will affect the sorting result of the product.

Step 3: Product popularity affects the sales volume of the product, and also affects the trust and recognition of consumers in the store to which the product belongs.

Step 4: The platform vane can affect the entire ranking and logistics, after-sales, as well as evaluation and customer service.

Step 5: Attribute filtering rules, use the attribute filtering function to further narrow the search intent, and bring rich, accurate and good experience to consumers.

Step 6: Criteria for judging cheating behaviors, including false transactions, misplacing categories and attributes, changing products, randomly filling in keywords in titles, etc.

Step 7: Repeated distribution. Products that are completely identical in the store and have completely identical important attributes are only allowed to be published once. Violation of the above rules will be considered as repeated distribution.