Extended Role Play Utterance Set Generation for Dialogue Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Role play-based dialogue systems face high costs and inefficiencies in collecting question answering data, and existing methods to enhance chat dialogue systems by converting sentences into character-specific speech may not adequately reflect the personality of the speaker.
Innovation Solution
An extended role play-based utterance set generation apparatus that combines role-played and non-role-played utterance sets to create extended utterance sets, allowing for increased variation and reflection of character personality at a lower cost, by selecting and associating non-role-played utterances with role-played ones based on similarity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If role play-based question answering data is collected from multiple workers, then collection efficiency and cost-effectiveness improve, but the total cost remains significant and collection time increases
Solution Approach 1:
The patent applies preliminary action by pre-collecting non-role-played utterance sets from multiple workers in advance, then reusing and combining these pre-collected datasets with role-played utterance sets. This eliminates the need to collect fresh data for each role-play scenario, significantly reducing collection time and computational resources while maintaining data diversity and quality.
Solution Approach 2:
The patent uses copying by creating extended role play-based utterance sets through association and combination of existing role-played utterance sets with non-role-played utterance sets. Instead of collecting entirely new data for each role-play scenario, the system copies and adapts existing utterance patterns, reducing both collection time and costs while preserving character personality representation.
2Adaptability or versatility
If sentence conversion to character-specific speech is used to enhance chat dialogue systems, then personality reflection improves, but the dialogue act and content may not adequately reflect the speaker's personality
Solution Approach 1:
The patent applies merging by combining role-played utterance sets (which contain character personality) with non-role-played utterance sets (which contain diverse dialogue acts and natural conversation patterns). The association unit merges these datasets to create extended role play-based utterance sets that simultaneously preserve character personality and maintain accurate, natural dialogue acts, resolving the contradiction between personality reflection and dialogue accuracy.
Solution Approach 2:
The patent uses composite materials analogy by creating composite utterance sets that integrate two different data sources: role-played utterances (providing personality characteristics) and non-role-played utterances (providing natural dialogue structure). This composite approach ensures both personality reflection and dialogue act accuracy are achieved simultaneously in the training data.
Data Source
AI summary
An extended role play-based utterance set generation apparatus includes a first data store storing role play-based utterance sets and a second data store storing non-role-played utterance sets. The role play-based utterance sets include a first query and a role play-based response to the query. The non-role-played utterance sets include a second query and a non-role-played response to the query. The disclosed technology determines similarity between the role play-based response and the non-role-played response. Upon determining that the role play-based response is the same or similar to the non-role-played response, the disclosed technology generates an association between the role play-based response and the second query and extends the role play-based utterance sets in the first data store with the second query.