Document Search Filtering by Word Frequency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional document search technologies using document vectors struggle to identify strong relationships between documents, often associating documents with weak relevance and overwhelming users with numerous irrelevant documents.

Innovation Solution

An information processing device that acquires documents and clue information, generates inter-document relationship source information, and filters documents based on high frequency appearance and vicinity appearance information to create an inter-document relationship table, thereby preventing association with weakly related documents.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If documents are associated with all words included in the document, then more documents are found, but documents with weak relationships are incorrectly associated

Engineering Contradiction:

Improvenumber of associated documentsVSAvoidrelationship strength

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

The patent changes the parameter of word frequency by filtering out words that appear in only one document. This parameter change ensures that only words with sufficient frequency (appearing in multiple documents) are used for association, thereby preventing weak relationships while maintaining meaningful connections between documents

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent applies local quality by treating different words differently based on their frequency characteristics. High-frequency words (those appearing in multiple documents) are used for association, while low-frequency words (appearing in only one document) are excluded. This selective approach ensures that association is based on meaningful, recurring patterns rather than isolated occurrences

Inventive Principle:

Principle #3Local quality

2Quantity of substance

If all words in a document are used for association, then comprehensive coverage is achieved, but user understanding of relevant documents becomes difficult

Engineering Contradiction:

Improvecoverage of associated documentsVSAvoiduser comprehension

Core Design Contradiction:

Quantity of substanceVSEase of operation

Solution Approach 1:

The patent extracts and removes words that appear in only one document from the association process. By taking out these low-frequency words, the system reduces the noise and irrelevant associations, making the remaining associations more meaningful and easier for users to comprehend

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the frequency parameter threshold for word inclusion in associations. By setting a minimum frequency requirement (words must appear in at least two documents), the system filters out irrelevant associations and presents a more manageable, comprehensible set of related documents to users

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12216699B2Information processing device, and generation method

Publication Date: 2025.02.04 MITSUBISHI ELECTRIC CORP

AI summary

An information processing device includes an acquisition unit that acquires a plurality of documents and clue information including a plurality of words and a generation processing unit that generates inter-document relationship source information based on the plurality of documents and the clue information, generates high frequency appearance information based on the inter-document relationship source information, generates vicinity appearance information based on the high frequency appearance information, and generates an inter-document relationship table based on the vicinity appearance information.