Information Delivery via Rank-Based User Interest Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional information delivery methods prioritize certain types of information but often deliver them to users who are not interested, resulting in a poor information delivery effect due to random selection without considering user preferences.
Innovation Solution
An information delivery method and device that determine a candidate information list based on user requests, predict ranks of information using a prediction model, and select designated delivery information based on current and historical average ranks to ensure targeted delivery matching user interests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the information platform configures a proportion of dedicated flow to deliver supported type of information randomly, then the supported type of information is delivered with priority, but the information may be delivered to users who are not interested in it, resulting in poor information delivery effect
Solution Approach 1:
The system establishes a feedback mechanism by analyzing user behavior data (clicks, views,停留时间) to continuously optimize information delivery decisions. The platform uses this feedback to adjust which supported information is delivered to which users, transforming the random selection into a data-driven decision process that maintains delivery priority while improving delivery effectiveness.
Solution Approach 2:
The system changes the parameters of information delivery by introducing user preference parameters and information quality parameters into the delivery decision process. Instead of random selection within the dedicated flow proportion, the system dynamically adjusts delivery based on matching user interests with supported information characteristics, thereby maintaining priority delivery while improving delivery effect.
2Reliability
If the information platform analyzes user requests to determine user preference, then the user may obtain more useful information, but the platform needs to balance user needs, support requirements, and producer interests
Solution Approach 1:
The system segments the information delivery decision into multiple independent modules: user preference analysis module, supported information identification module, producer interest consideration module, and final delivery decision module. Each module handles a specific aspect of the balancing act, reducing overall system complexity while maintaining comprehensive consideration of all factors.
Solution Approach 2:
The system introduces an intermediary scoring mechanism that translates user preferences, supported information requirements, and producer interests into a unified evaluation score. This intermediary layer simplifies the balancing process by converting multiple conflicting requirements into a single decision criterion, reducing the complexity of making balanced delivery decisions.
Data Source
AI summary
An information delivery method is provided, including: determining a candidate information list in response to a page request sent by a user, where the candidate information list comprises a plurality pieces of information; generating ranks of the pieces of information in the candidate information list by sorting the plurality pieces of information in the candidate information list in accordance with a predetermined rule; acquiring a current average rank of the information of a designated type in the candidate information list based on the ranks of the pieces of information in the candidate information; and selecting designated delivery information from the information of the designated type, based on the current average rank of the information of the designated type and a historical average rank corresponding to the information of the designated type.