Dynamic sorting adjustment method based on exposure

Abstract

The invention discloses a dynamic sorting adjustment method based on exposure, and relates to the technical field of electronic commerce commodity display. The system comprises a commodity library, a user library, a weight module, a sorting module, a weight adjusting module, a weight limiting module, a stay duration acquisition module and an exposure calculation module. The weight limiting module is used for manually limiting the weight of the corresponding commodity or shop; the weight module interacts with the commodity library and the user library; the staying duration acquisition module is used for acquiring the duration data of the commodities or shops which are sequenced and displayed by the sequencing module under the current weight and browsed by the user in real time, and counting the exposure of the corresponding browsed commodities or shops based on the quantity of the commodities or shops corresponding to all pages displayed at the time; and the data value of the exposure is fed back to the weight adjusting module for automatic adjustment. According to the invention, the labor cost of commodity sorting is reduced; optimization is achieved aiming at sorting of the exposure rates, and continuous optimization is achieved aiming at user behaviors.

Description

technical field [0001] The invention belongs to the technical field of e-commerce commodity display, and in particular relates to a dynamic ranking adjustment method based on exposure. Background technique [0002] The current sorting algorithm for product search is mainly based on the transaction volume and positive rate to perform limited-level sorting of product searches, and at the same time assists some manual intervention in the sorting results, and manages risk control and overall data arrangement for some abnormal data. [0003] For a large number of existing e-commerce systems, two solutions are often adopted according to the size of the system. In the case of a small size, strong manual intervention is used to continuously maintain a small amount of commodity data. After the data volume reaches the upper limit of manual maintenance capabilities , adopts the sorting based on simple data such as the number of married couples and positive ratings, and also inserts a s...

AI Technical Summary

Problems solved by technology

[0004] The existing product search and sorting schemes have the following disadvantages: (1) The labor cost of maintaining the product sorting: the existing lightweight system requires manpower to directly maintain the products, and the frequent operations of product loading and unloading and product sorting require continuous maintenance by dedicated personnel , requiring a large amount of manpower; (2) Defects based on transaction volume and favorable rate: platforms based on transaction volume and favorable rate generally have already begun to take shape. , so it is difficult to avoid malicious swiping among merchants. If things go on like this, it will cause swiping and well-known brands to occupy the first place in the product search structure for a long time, crowding out the living space of small merchants, and eventually causing the ecological imbalance of the platform; (3) Lack of feedback on user behavior : The traditional sorting method can only sort and optimize the user's purchase and evaluation behaviors, but the user's behavior is complex and diverse. The most common operations users do are browsing and clicking. These data cannot be entered into the system, nor can they be processed. The basis for tuning the training logic, resulting in data loss

Images