Keyword-Advertising Match Optimization via Linear Programming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online trading platforms often fail to ensure that all product areas offered by sellers are adequately represented in advertising matches, leading to potential buyers not finding relevant advertising sources and degrading server performance due to limited keyword associations and prioritization of the latest advertising sources.
Innovation Solution
A method using linear programming to optimize the match between advertising sources and keywords, ensuring that each product area is covered by selecting advertising sources based on categories and generating keyword sets to maximize seller gain, thereby improving retrieval performance and user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If advertising sources are matched to keywords based on latest advertising sources or maximum gain for current advertising source, then the matching speed and simplicity are improved, but the coverage of all product areas offered by the seller deteriorates
Solution Approach 1:
The patent segments the matching process into two distinct phases: (1) generating multiple candidate matching results using different criteria (latest advertising sources, maximum gain for current advertising source), and (2) selecting the optimal matching result by evaluating product area coverage. This segmentation allows each phase to optimize for its specific goal without compromising the other.
Solution Approach 2:
The patent generates multiple candidate matching results (excessive action) beyond what a single criterion would produce, then selects the appropriate one. By generating more candidates than necessary and filtering them, the system ensures comprehensive product area coverage while maintaining matching efficiency.
2Device complexity
If advertising sources are matched to keywords based on latest advertising sources or maximum gain for current advertising source, then the matching process is simplified, but the seller's total gain deteriorates
Solution Approach 1:
The patent generates multiple candidate matching results using simple, fast criteria (excessive action), then applies a selection mechanism to choose the one that maximizes seller's total gain. This approach keeps the generation process simple while ensuring optimal gain through selective evaluation.
Solution Approach 2:
The patent incorporates a feedback mechanism where the system evaluates the total gain resulting from each candidate matching result and uses this information to select the optimal matching. The feedback loop ensures that the matching process continuously optimizes for seller gain without requiring complex real-time calculations.
3Productivity
If the match between advertising sources and keywords is generated based on latest advertising sources or maximum gain for current advertising source, then the retrieval performance of the online trading platform server deteriorates due to repeated searches
Solution Approach 1:
The patent performs preliminary action by pre-generating multiple candidate matching results and selecting the optimal one before actual keyword queries are processed. This preliminary preparation ensures that when buyers search, the system can quickly retrieve pre-optimized matching results, reducing repeated searches and improving retrieval performance.
Solution Approach 2:
The patent segments the retrieval process into offline candidate generation and online selection/execation. The computationally intensive candidate generation happens offline, while the online phase quickly selects and executes the pre-determined optimal matching, significantly reducing retrieval time and server load during peak usage.
Data Source
AI summary
Providing online advertisements includes selecting, among a plurality of advertising sources provided by a seller, a selected plurality of advertising sources that meet a predefined condition; generating a plurality of keyword sets that correspond to the selected plurality of advertising sources; establishing a programming model according to a set of predefined constraints, wherein the programming model represents match of the selected plurality of advertising sources and the plurality of keyword sets; and determining a substantially optimal match between at least some of the plurality of advertising sources and the plurality of keyword sets by solving the programming model.