Speech Consumability Assessment via Sentence Length Variation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional readability tests for texts are inadequate in assessing the understandability and engagement of audio products generated by text-to-speech tools, as they primarily focus on sentence and syllable counts without considering the variability in sentence length and user experience.
Innovation Solution
A method and system that calculate a consumability-readability score by determining the mean sentence length and variations in sentence lengths, using a text matrix to evaluate the input text and produce an output speech that is engaging and easy to understand, incorporating natural language processing to adjust the text if necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional readability tests based on sentence count, word count, and syllable count are used, then the understandability of text can be measured, but the engagement and attentiveness of listeners during audio playback cannot be assessed
Solution Approach 1:
The patent transforms conventional readability parameters (sentence count, word count, syllable count) into a new parameter - average sentence length variation - that specifically measures audio engagement. By calculating the absolute differences between each sentence length and the mean sentence length, then averaging these variations, the system creates a metric that reflects listener attentiveness during audio playback, thus adapting the measurement to the audio context while maintaining understandability assessment.
2Measurement precision
If text is optimized for reading comprehension, then understandability improves, but the text may become boring and less engaging when converted to speech
Solution Approach 1:
The patent implements a feedback mechanism where the average sentence length variation score is calculated from the input text and provided back to users. This feedback allows users to understand how their text will perform in audio format and make adjustments before production. By showing the variation score and allowing text modification to improve it, the system enables users to optimize their content for both understandability and audio engagement without requiring complex post-production adjustments.
Solution Approach 2:
The patent performs preliminary analysis of the input text to calculate average sentence length variation before audio production. This preliminary action identifies potential engagement issues in the text structure, allowing users to revise the text in advance. By detecting and flagging texts with excessive sentence length uniformity before conversion to speech, the system prevents the creation of boring audio content, thus improving audio production quality proactively rather than reactively.
3Device complexity
If sentence lengths are made uniform for ease of processing, then text processing becomes simpler, but listener attentiveness and engagement decrease
Solution Approach 1:
The patent introduces dynamic variation in sentence lengths as a desirable characteristic for audio content. Rather than seeking uniformity, the system calculates and rewards variation by measuring the absolute differences between each sentence length and the mean. This dynamic approach encourages diverse sentence structures that maintain listener attentiveness during audio playback, transforming the static requirement for uniform processing into a dynamic criterion for engagement optimization.
Data Source
AI summary
Methods, computer program products, and systems are presented. The methods include, for instance: obtaining an input text for an output speech. The number of words and syllables are counted in each sentence, and a mean sentence length of the input text is calculated. Each sentence length is checked against the mean sentence length and a variation for each sentence is calculated. For the input text, the consumability-readability score is produced as an average of variations for all sentences in the input text. The consumability-readability score indicates the level of satisfaction for the listener of the output speech based on the input text.