Search Query Synonym Generation via Session Log Filtering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods for generating synonyms for search queries in web search engines are inadequate, as they fail to recognize specific brand names, product names, and retail-specific jargon, and are not effective for tail queries due to lower online traffic and limited information, leading to irrelevant search results.

Innovation Solution

A system and method that utilizes session log reformulation to analyze query data, categorize search queries, and filter synonym candidates to generate an accurate set of synonyms by identifying and filtering synonym pairs based on frequency and category matching, ensuring relevant search results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If out-of-the-box ontology technologies are used to generate synonyms, then the system can generate synonyms for known products and services, but it fails to recognize specific brand names, product names, and retail-specific jargon

Engineering Contradiction:

Improvesynonym generation capabilityVSAvoidrecognition accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent segments the synonym generation process into multiple independent modules: ontology-based synonym generation, query log analysis, and session log reformulation analysis. Each module handles specific types of queries independently, allowing the system to maintain high reliability for different query types while preserving overall adaptability.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces query log analysis and session log reformulation as intermediary processes between the user query and the final search results. These intermediaries capture real-world usage patterns and reformulations, bridging the gap between generic ontology synonyms and specific brand/product recognition.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If query logs are analyzed for reformulation to generate synonyms, then head queries can obtain suitable synonyms, but tail queries with lower online traffic produce less effective synonyms

Engineering Contradiction:

Improvesynonym generation efficiencyVSAvoidsynonym accuracy for tail queries

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary action by pre-processing and storing session log reformulations in an offline manner before they are needed for actual search queries. This allows tail queries to benefit from pre-computed synonyms even when online traffic is low, improving synonym accuracy without affecting real-time search productivity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates copies of reformulated queries from session logs and stores them as synonym candidates. This copying mechanism allows the system to leverage patterns from head queries and apply them to tail queries, ensuring consistent synonym quality across different query frequency levels.

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If multiple synonym generation methods are combined, then comprehensive synonym coverage is achieved, but system complexity increases

Engineering Contradiction:

Improvesynonym coverageVSAvoidsystem architecture

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent divides the complex synonym generation system into three separate, independently deployable modules: ontology-based generation, query log analysis, and session log reformulation analysis. Each module can be developed, maintained, and optimized independently, reducing overall system complexity while maintaining comprehensive synonym coverage.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent designs a universal framework where multiple synonym generation methods share common infrastructure for processing and integrating results. This multi-functional architecture allows the system to handle different query types uniformly while leveraging the strengths of each individual method, achieving comprehensive coverage without proportionally increasing complexity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS11829395B2System, method, and non-transitory computer-readable storage media for generating synonyms of a search query

Publication Date: 2023.11.28 WALMART APOLLO LLC

AI summary

A method including analyzing the query data to formulate one or more groups of synonym candidates, wherein each synonym in the one or more groups of synonym candidates is categorized in one or more synonym candidate categories. Reformulating the first search query by removing at least one synonym candidate from the first group of synonym candidates to form a first set of synonyms such that each synonym in the first set of synonyms is categorized in the synonym candidate category that matches the search query category of the search query, wherein the at least one synonym candidate removed from the first group of synonym candidates is not categorized in the synonym candidate category.