Query Similarity Analysis via Time Series Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advertisers face challenges in maximizing the effectiveness of their advertising dollars as they struggle to accurately predict query trends and identify similar queries to optimize their bid terms and advertisement placement.
Innovation Solution
A query analysis system represents each query as a time series of frequencies and generates overall and partial similarity scores using metrics like cosine similarity and Euclidean distance to identify similar queries, allowing for the selection of queries with the highest similarity scores for targeted advertising.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If advertisers use traditional query trend analysis methods, then they can identify basic query patterns, but they cannot accurately predict future query frequencies or identify truly similar queries for optimal advertising placement
Solution Approach 1:
The patent segments the query analysis process into distinct components: (1) receiving query frequencies over time intervals, (2) generating overall similarity scores based on complete time series, and (3) generating partial similarity scores based on segmented time sub-series. This segmentation allows the system to handle complex analysis through manageable stages, improving prediction accuracy while maintaining systematic complexity control
Solution Approach 2:
The patent transforms the one-dimensional query frequency data into a multi-dimensional analysis by creating both overall similarity scores (comparing complete time series) and partial similarity scores (comparing segmented time sub-series). This dimensional expansion enables more nuanced query similarity assessment and trend prediction, resolving the contradiction between accuracy and complexity
2Reliability
If the system calculates overall similarity scores for all queries, then comprehensive query similarity can be identified, but computational resources and time are excessively consumed
Solution Approach 1:
The patent implements partial action by calculating not only overall similarity scores for complete time series but also partial similarity scores for specific time sub-series. This allows the system to identify query similarities at different granularities, providing reliable results for both comprehensive and targeted analysis scenarios while optimizing computational resource utilization
3Productivity
If advertisers increase bid amounts for predicted high-frequency queries, then advertising effectiveness may improve, but advertising costs increase significantly
Solution Approach 1:
The patent establishes a feedback mechanism where query similarity analysis and trend prediction results directly inform advertising bid decisions. By identifying queries with high overall and partial similarity scores to target queries, the system provides actionable intelligence that enables advertisers to optimize bid amounts based on predicted performance, improving advertising effectiveness while controlling budget consumption through data-driven decision-making
Data Source
AI summary
Techniques for identifying similar queries based on their overall similarity and partial similarity of time series of frequencies of the queries are provided. To identify queries that are similar to a target query, the query analysis system generates, for each query, an overall similarity score for that query and the target query based on the time series of the query and the target query. The query analysis system also generates, for each query, partial similarity scores for the query and the target query based on various time sub-series of the overall time series of the queries. The query analysis system then identifies queries as being similar to the target query based on the overall similarity scores and the partial similarity scores of the queries.