Natural Language Call Router Adaptation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing natural language call routers are inefficient in adapting to rapid changes in caller inputs due to complex modifications required in statistical language models and classifiers, which involve extensive manual intervention and long compilation times.

Innovation Solution

A natural language call router utilizing a robust parser and interpreter that detects classes of words with common attributes, allowing for easy modification and recompilation, using a class-based statistical language model and rule-based destination determination, reducing the need for manual retagging and retraining.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a Statistical Semantic Model (SSM) is used as the classifier, then the system can understand statistical patterns of human word usage, but changing the classifier requires manual retagging of tens of thousands of transcribed utterances and retraining that takes many hours of compilation time

Engineering Contradiction:

ImproveAbility to adapt to changing caller inputsVSAvoidCompilation time for retraining the classifier

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent segments the classifier into multiple specialized parsers (robust parser, statistical parser, hybrid parser), each handling specific types of linguistic patterns. This segmentation allows individual parsers to be modified and retrained independently rather than requiring complete system retraining, significantly reducing compilation time when adapting to new caller inputs

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system dynamically selects and combines different parsing strategies based on the input characteristics. The robust parser provides deterministic rule-based matching, the statistical parser handles probabilistic patterns, and the hybrid parser combines both approaches. This dynamic adaptation allows the system to respond flexibly to changing inputs without requiring full retraining

Inventive Principle:

Principle #15Dynamics

2Ease of manufacture

If a program or robust parsing grammar is used as the classifier, then the system can route calls based on keyword detection, but changing the classifier involves complicated modifications with many possible word sequences

Engineering Contradiction:

ImproveEase of modifying the classifierVSAvoidComplexity of handling multiple word sequences

Core Design Contradiction:

Ease of manufactureVSDevice complexity

Solution Approach 1:

The patent creates a universal parsing framework that handles multiple word sequences and linguistic patterns through a single unified architecture. The robust parser uses parameterized templates and wildcards that can match various word sequences without requiring separate programs for each pattern, simplifying modifications while maintaining comprehensive coverage

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS8942981B2Natural language call router

Publication Date: 2015.01.27 CELLCO PARTNERSHIP INC

AI summary

A natural language call router forwards an incoming call from a caller to an appropriate destination. The call router has a speech recognition mechanism responsive to words spoken by a caller for producing recognized text corresponding to the spoken words. A robust parsing mechanism is responsive to the recognized text for detecting a class of words in the recognized text. The class is defined as a group of words having a common attribute. An interpreting mechanism is responsive to the detected class for determining the appropriate destination for routing the call.