Search Term Rewriting via Historical Data Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing search technologies face inaccuracies and low recall rates due to rule-based rewriting of search terms, leading to suboptimal search results.

Innovation Solution

A method and apparatus that utilize a database to match and rewrite search terms based on historical search data, determining optimal search term candidates by processing properties and thresholds, thereby eliminating the reliance on fixed rules and improving search accuracy and recall.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If rule-based rewriting of search terms is used, then search coverage is improved, but search accuracy deteriorates

Engineering Contradiction:
Improvesearch coverageVSAvoidsearch accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent implements feedback mechanisms by analyzing user behavior data (click-through rates, dwell time, conversion rates) from search results and using this feedback to iteratively optimize the rewriting rules. The system continuously learns from user interactions to improve both coverage and accuracy of search term rewriting, resolving the contradiction between broader coverage and maintained accuracy.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent dynamically adjusts rewriting parameters such as rewriting probability, rule priority weights, and threshold values based on contextual factors including query type, user profile, and historical performance data. This allows the system to adapt the degree and nature of rewriting to maintain accuracy while maximizing coverage across different search scenarios.

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If simple rule-based rewriting is used, then system complexity is reduced, but rewriting accuracy deteriorates

Engineering Contradiction:
Improvesystem complexityVSAvoidrewriting accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent introduces intermediary components including a rule engine that mediates between simple rewriting rules and complex decision-making, a context analyzer that processes query characteristics, and a performance evaluator that assesses rewriting quality. These intermediaries enable accurate rewriting without requiring the entire system to be complex, modularizing the complexity management.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the rewriting system into independent modular components: rule generation module, rule execution module, context analysis module, and performance evaluation module. Each module handles specific aspects of the rewriting process, allowing simple rules to be combined with sophisticated analysis through clear separation of concerns, maintaining manageability while achieving accuracy.

Inventive Principle:
Principle #1Segmentation

3Adaptability or versatility

If extensive rule-based rewriting is used, then search term coverage is improved, but error rate increases

Engineering Contradiction:
Improvesearch term coverageVSAvoiderror rate
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent applies partial rewriting by selectively rewriting only portions of search terms that benefit from transformation, rather than applying extensive rewriting to all terms. The system uses rewriting probability thresholds and contextual filters to determine when rewriting should be applied, avoiding excessive rewriting that would increase errors while still achieving adequate coverage through targeted transformations.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent implements validation and verification mechanisms before rewriting is applied, including context checks, similarity thresholds, and confidence score calculations. These preliminary checks cushion against potential errors by identifying cases where rewriting should not be applied, preventing error propagation while maintaining coverage through careful selection of rewriting candidates.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentUS9576054B2Search method, apparatus and system based on rewritten search term
Publication Date: 2017.02.21 ALIBABA GROUP HOLDING LTD
  • US9576054B2 patent drawing
  • US9576054B2 patent drawing
  • US9576054B2 patent drawing

AI summary

A search method is disclosed. In the method, a data rewriting system obtains, from a database, one or more search term candidates that are relevant to a present search term. The data rewriting system retrieves properties of the present search term and the one or more search term candidates, where the properties describe respective matching results of the present search term and the one or more search term candidates. Based on the matching results, the data rewriting system determines whether the present search term needs to be rewritten, and rewrites the present search term based on the matching results to provide a rewritten present search term in response to determining that the present search term needs to be rewritten. A search engine performs a search based on the rewritten present search term.