Metadata Text Retrieval TRIZ Cases
- Document Bias Context Rendering System
System identifies biased document portions and renders related context resources, reducing computational load by pre-computing source analysis.
- Personalized Content Recommendation System Using Segmented Data Processing
A system processes user behavior data to generate personalized content recommendations for website visitors.
- Digital Document Recognition via Text Search and Feature Comparison
A digital document recognition system extracts textual content to search an indexed master database and generates a candidate list for precise feature comparison.
- Comment Opinion Classification via Automatic Training Data Generation
A classification model identifies opinion categories in comments using automatically generated training samples from a comment dictionary.
- Document Retrieval System Using Classification Code Definition Sentences
A document retrieval program extracts candidate classification codes from database documents to identify relevant metadata for search queries.
- Author-Created Digital Agents for Document Content Corpora
Author-created digital agents search curated content corpora to deliver relevant information directly within electronic documents.
- ASR Confidence Score Calibration for Transcription Accuracy
Calibrating ASR confidence scores via Hidden Markov models corrects transcription errors from untrained users without external rule definitions.
- Metadata-Driven Data Reclassification Without Re-Access
Associates metadata with stored objects to generate classification recommendations, reducing time consumption when business rules change.
- Clustering Engine for Unsupervised Document Tagging
An automated clustering engine processes customer support inquiries to reduce response times while managing system complexity through self-service routing.
- Document Attribute Search Indexing via Manifest Templates
Explicit document attribute specification through publisher manifests improves search accuracy and reduces resource usage by eliminating inference errors.
- Distinct Author Identification System for Bibliographic Disambiguation
A distinct author identification system groups publications into entity clusters using unique codes and similarity metrics.
- Voice Retrieval Output Redundancy Elimination
A retrieval device generates voice output by erasing redundant item values based on commonalities between words and facility types.