Utterance Feature Word Extraction for Vehicle Information Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information providing systems struggle to accurately determine the purpose of actions of vehicle occupants without recognizing specific conversation contents, leading to ineffective information provision.
Innovation Solution
An information providing device and system that acquires and analyzes utterance data to extract feature words, count their occurrences at specific locations or destinations, and provide information based on these counts, without requiring specific conversation content recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If specific conversation contents are recognized to determine the purpose of actions, then information provision accuracy is improved, but privacy protection deteriorates and system complexity increases
Solution Approach 1:
The patent extracts only the essential feature words from conversation contents rather than processing the entire conversation text. The utterance content acquisition unit collects conversation data, the feature word extraction unit identifies key terms, and the utterance number counting unit counts occurrences of these feature words at different locations. This extraction approach enables purpose determination while protecting privacy by not transmitting or storing specific conversation contents.
Solution Approach 2:
The patent segments conversation analysis into distinct components: feature word extraction, location-based counting, and purpose determination. By dividing the utterance content into individual feature words and analyzing their distribution across different locations (road links, crossings, areas, POIs), the system achieves accurate purpose determination without handling complete conversation texts, thus balancing accuracy with privacy protection.
2Measurement precision
If complete conversation contents are transmitted for analysis, then analysis accuracy is improved, but data transmission volume and processing complexity increase
Solution Approach 1:
The feature word extraction unit isolates only the necessary keywords from complete conversation contents. Instead of transmitting entire conversation texts to the server, only these extracted feature words are transmitted along with location information. This significantly reduces data transmission volume and processing complexity while maintaining the ability to accurately determine occupant purposes.
Solution Approach 2:
The feature word extraction is performed in advance at the vehicle terminal before data transmission to the server. This preliminary processing step prepares the data in a condensed format (feature words with location tags) that requires minimal further processing at the server end, thereby reducing overall system complexity and transmission requirements.
3Measurement precision
If feature words are counted for all locations, then movement intention detection accuracy is improved, but data processing time and computational load increase
Solution Approach 1:
The patent applies local quality by focusing counting operations on specific location types relevant to movement intentions (road links, crossings, areas, POIs) rather than uniformly processing all possible locations. The utterance number counting unit counts feature word occurrences specifically at these defined location indicators, enabling efficient movement intention detection by concentrating computational resources on geographically relevant data points.
Data Source
AI summary
An information providing device includes: an utterance contents acquiring unit configured to acquire contents of utterance; a feature word extracting unit configured to extract a feature word; an utterance-time information acquiring unit configured to acquire information on a position or a destination of the vehicle when the feature word has been uttered; an utterance number counting unit configured to count the number of times of utterance of each of a plurality of different words for each of position indicators; a current information acquiring unit configured to acquire information on a position or a destination of a predetermined target vehicle; and an information providing unit configured to provide information to the user based on the counting results of the position indicator corresponding to the position of the target vehicle or the destination indicator corresponding to the destination of the target vehicle from the utterance number counting unit.