Knowledge Representation TRIZ Cases
- Graph Embedding Model for Node Classification Accuracy
A neural network graph embedding model generates numerical node representations using combined graph and non-graph data.
- Neural Architecture Search via Dynamic Block Capacity Adjustment
A neural architecture search method adjusts student network block capacities based on teacher-student feature map loss functions.
- Concurrent Content Filtering Apparatus for Streaming Data
A content filtering apparatus parses data containers into components and directs each component to a respective filtering module.
- Automated Coverage Assistant for Customer Service Gap Detection
An automated coverage assistant module identifies gaps in customer service coverage by analyzing current FAQs, IVR scripts, and agent scripts.
- Programmable Packet Processing Graph for Dynamic Network Flow Management
Segmenting control logic from switching logic resolves the contradiction between device complexity and network adaptability.
- Knowledge Anchored AI Tree for Explainable Ecommerce Automation
A knowledge anchored artificial intelligence system uses a traversable tree structure to process queries and execute actions.
- Social Network Member Engagement Prediction
A system ranks potential members to increase interactions within explicit social networks.
- Modular Message Queue for Ocean of Things Bandwidth Optimization
Priority-based message packing and dynamic queue management transmit critical sensor data while preventing bandwidth saturation in Ocean of Things networks.
- Sequence-Based Compliance Measurement for Business Process Models
A sequence-based compliance measurement system analyzes firing sequences to evaluate business process adherence against reference models.
- Classifier Training with Knowledge Graph Embeddings
A classifier training method uses knowledge graph embeddings to guide machine learning systems.
- Dispute Resolution Procedure Selection System
A dispute resolution procedure selection system analyzes party characteristics and goals to recommend appropriate processes.
- RNN Term Correlation Mining for Large-Scale Documents
LSTM model extracts term correlations from large-scale documents using co-occurrence analysis.
- Data Quality Tool Dynamic Rule Generation
A data quality tool employs machine learning to generate dynamic rules from metadata and statistical properties.