Presentation Assistance Device for Forbidden Word Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
During presentations, speakers may inadvertently use forbidden words, and existing systems do not effectively alert or correct them in real-time to maintain a smooth explanation.
Innovation Solution
A presentation assistance device that stores keywords and related 'no-good' words, uses speech recognition to identify spoken terms, and outputs alerts when forbidden words are spoken, while displaying relevant keywords to guide the speaker in correcting their statements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If speech recognition and real-time monitoring are implemented during presentations, then the ability to detect and correct forbidden words is improved, but the device complexity and processing requirements increase
Solution Approach 1:
The system segments the presentation monitoring function into distinct modules: speech recognition means for converting speech to text, term determination means for comparing recognized terms against stored keywords and NG words, and separate output means for displaying keywords versus displaying alerts. This modular segmentation allows each component to perform its specific function efficiently while maintaining overall system reliability without excessive complexity.
Solution Approach 2:
The term determination means acts as an intermediary component that bridges the speech recognition means and the output means. It receives recognized terms, compares them against stored keyword and NG word databases, determines whether each term is a keyword or NG word, and directs the appropriate output. This intermediary layer simplifies the overall system architecture by centralizing the decision-making logic.
2Loss of information
If multiple storage components for keywords, related words, and NG words are implemented, then the comprehensiveness of speech monitoring is improved, but the information processing load and system complexity increase
Solution Approach 1:
The system applies local quality by storing different types of words (keywords, related words, NG words) in separate storage means with distinct purposes. Keywords are stored for positive reinforcement and guidance, while NG words are stored for negative feedback and correction. This separation allows each storage component to be optimized for its specific function, improving information completeness without requiring a single complex unified database.
Solution Approach 2:
Instead of storing all words in a single database and filtering them during processing, the system inverts the approach by pre-separating words into categories (keywords and NG words) in separate storage means before processing. This inversion simplifies the term determination process, as the system only needs to check against appropriate storage means based on whether the recognized term should be encouraged or corrected.
3Loss of time
If real-time speech recognition and term determination are performed, then the ability to provide immediate feedback is improved, but the processing time and computational resources increase
Solution Approach 1:
The system performs preliminary action by pre-storing keywords and NG words in separate storage means before the presentation begins. During the presentation, the term determination means only needs to compare recognized terms against these pre-prepared lists rather than performing complex analysis in real-time. This preliminary preparation significantly reduces processing time and energy consumption during the actual presentation while maintaining immediate feedback capability.
Data Source
AI summary
To provide a presentation assistance device that can display keywords related to presentation materials and call attention by displaying an alert when words that are forbidden to speak are spoken. A presentation assistance device 1 comprises: a presentation material storage means 3; a keyword storage means 5 which stores a plurality of keywords related to presentation materials; a related word storage means 7 which stores one or a plurality of related words for each of the plurality of keywords; an NG word storage means 9 which stores one or a plurality of NG words for each of the plurality of keywords; a voice recognition means 11; a term determination means 15 which determines whether a voice recognition term corresponds to a related word or an NG word; and a keyword output means 17 which when the voice recognition is a related word, outputs a keyword related to the related word, and when the voice recognition term is an NG word, outputs an alert and a keyword related to the NG word.