Keyword Mapping for Ad Recall via Synonym Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The recall rate of promotion information is low in existing advertisement promotion services due to the difficulty in matching retrieval entries with binding keywords, especially when synonyms or related terms are used, leading to ineffective keyword hits.
Innovation Solution
An information processing method and apparatus that generates first-category and second-category mapping data by matching stored keywords with historical search keywords, using document vector models to calculate similarities and extend keyword mappings, allowing for more accurate matching of target keywords with stored keywords, even with a smaller number of initial binding keywords.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a bidding rank mechanism is used with multiple binding keywords, then advertisement promotion effectiveness is improved, but the number of keywords an advertiser can set is limited due to cost constraints
Solution Approach 1:
The system pre-generates extended keywords from historical search records and synonyms before actual advertisement matching occurs. This preliminary expansion of the keyword pool allows advertisers to effectively cover more search terms without increasing their direct keyword purchase costs, resolving the contradiction between promotion effectiveness and keyword quantity constraints
Solution Approach 2:
The patent introduces an intermediary mechanism (synonym expansion module and historical search record analysis) that bridges the gap between limited paid keywords and broader search coverage. This intermediary layer translates user searches into matched advertisements even when exact keyword matches don't exist, maintaining promotion effectiveness without requiring proportional increases in keyword quantity
2Measurement precision
If exact keyword matching is used between binding keywords and retrieval entries, then matching precision is improved, but recall rate deteriorates when synonyms or related terms are used
Solution Approach 1:
The patent segments the keyword matching process into two distinct stages: exact matching for precision and synonym/extension matching for recall. The system first performs precise exact matching on paid keywords, then separately processes extended keywords from historical records and synonyms. This segmentation allows both exact matching precision and synonym recall to coexist without interfering with each other
Solution Approach 2:
The system dynamically adjusts the matching strategy based on the type of keyword and search context. For paid binding keywords, exact matching is applied to maintain precision. For extended keywords from historical records and synonyms, flexible matching is used to improve recall. This dynamic approach allows the system to optimize for precision or recall depending on the specific matching scenario
3Loss of energy
If the number of binding keywords is reduced to lower costs, then advertising cost is reduced, but the ability to match retrieval entries deteriorates
Solution Approach 1:
The system enables self-service keyword expansion by automatically analyzing historical search records and generating relevant extended keywords and synonyms. This self-service mechanism compensates for the reduced number of paid keywords by autonomously creating additional matching opportunities, maintaining keyword matching ability while reducing advertising costs
Solution Approach 2:
The patent changes the parameters of keyword matching by introducing extended keywords with different matching thresholds and weights. Instead of relying solely on exact matches of paid keywords, the system transforms the matching parameters to include fuzzy matching, synonym matching, and historical record-based matching, thereby maintaining matching ability with fewer paid keywords
Data Source
AI summary
A method of generating content information that matches at least one stored keyword is described. At least one keyword associated with content information is stored. At least one previously searched keyword in a search record is matched with the at least one stored keyword associated with the content information. First-category mapping data is generated based on a first mapping between the matched at least one stored keyword and the at least one previously searched keyword. Second-category mapping data is generated based on the content information and the at least one stored keyword. A received target keyword is determined to be included in the first-category mapping data. In response to the received target keyword, which is included in the first-category mapping data, circuitry of a terminal searches for the content information associated with the target keyword in the second-category mapping data and displays the content information.