Concept Graph Query Expansion for Relevant Ad Bidding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional advertising systems rely on keywords, leading to irrelevant ad displays when queries use long tail words or unrelated concepts, and ad buyers have limited control over bid relevance and efficiency due to fixed bid amounts per keyword.
Innovation Solution
Ad buyers use concept-based bid calculations, specifying bid functions that incorporate semantic and environmental information, allowing for dynamic ad content and targeting based on interpreted expressions, with the ability to modify bid functions and provide advertisements as questions or audio cues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If keyword-based advertising is used, then ad buyers can specify bid amounts on keywords, but queries using long tail words or unrelated concepts will result in irrelevant ad displays
Solution Approach 1:
The patent changes the fundamental parameter of ad targeting from keyword matching to concept-based matching. Instead of comparing exact keywords, the system extracts concepts from both queries and ad descriptions, then calculates bid amounts based on concept relevance. This allows queries with long tail words or synonymous expressions to match relevant ads while maintaining manageable complexity through automated concept extraction and scoring.
Solution Approach 2:
The patent introduces concepts as an intermediary layer between keywords and ad relevance determination. Rather than directly matching keywords to ads, the system extracts concepts from queries, compares them with concepts in ad descriptions, and uses this concept-level matching to determine bid amounts. This intermediary approach resolves the contradiction by enabling flexible, relevant ad targeting without requiring complex keyword enumeration.
2Ease of operation
If fixed bid amounts per keyword are used, then ad buyers have simple bidding control, but they have limited ability to optimize relevance and spending efficiency
Solution Approach 1:
The patent transforms static fixed bid amounts into dynamic bid calculations based on concept relevance. The bid amount is no longer a fixed value per keyword but is dynamically determined by the relevance score between query concepts and ad description concepts. This dynamic approach maintains ease of operation through automated calculations while dramatically improving ad campaign efficiency by ensuring bids reflect actual relevance.
Solution Approach 2:
The patent implements feedback mechanisms where the system continuously evaluates concept relevance between queries and ads, then adjusts bid amounts accordingly. The bid calculation incorporates feedback from concept matching quality, allowing ad buyers to automatically optimize spending efficiency based on real-time relevance assessment without manual intervention for each query.
3Measurement precision
If ad buyers specify bids on specific keywords, then they have precise control over which keywords trigger ads, but queries expressing relevant concepts with different wording will not hit the ad
Solution Approach 1:
The patent makes the ad targeting system universal by enabling it to recognize multiple expressions of the same concept. Instead of requiring separate keyword bids for each possible wording, the system extracts underlying concepts from queries and matches them with concepts in ad descriptions. This allows a single ad to be triggered by multiple different wordings and expressions of the same underlying concept, achieving both precision and versatility.
Solution Approach 2:
The patent extracts the essential meaning from queries by identifying underlying concepts rather than matching surface-level keywords. By taking out the core concept from various wordings and expressions, the system can match queries to ads based on semantic meaning rather than exact keyword matches. This extraction approach maintains precision in identifying relevant ads while expanding versatility to cover diverse query expressions.
Data Source
AI summary
Original concepts obtained from a query may be augmented with additional concepts connected to the original concepts in a concept graph in response to determining that the original concepts did not match a sufficient number of bid functions. The augmented set of concepts may then be evaluated with respect to the bid functions to identify matching ad functions. This process may be repeated until a sufficient number of matching ad functions are found. A bid amount of the matching bid functions may be calculated, such as based on semantic information obtained as a result of the query. The bid amounts may further be based on environmental information. A bid function is selected based on the bid amounts and the content associated with the bid function is provided to the source of the query. The content may be selected based on the semantic information.