ECLAIR: Enhanced Clarification for Interactive Responses
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910885872664576 |
|---|---|
| author | Murzaku, John Liu, Zifan Tanjim, Md Mehrab Muppala, Vaishnavi Chen, Xiang Li, Yunyao |
| author_facet | Murzaku, John Liu, Zifan Tanjim, Md Mehrab Muppala, Vaishnavi Chen, Xiang Li, Yunyao |
| contents | We present ECLAIR (Enhanced CLArification for Interactive Responses), a novel unified and end-to-end framework for interactive disambiguation in enterprise AI assistants. ECLAIR generates clarification questions for ambiguous user queries and resolves ambiguity based on the user's response.We introduce a generalized architecture capable of integrating ambiguity information from multiple downstream agents, enhancing context-awareness in resolving ambiguities and allowing enterprise specific definition of agents. We further define agents within our system that provide domain-specific grounding information. We conduct experiments comparing ECLAIR to few-shot prompting techniques and demonstrate ECLAIR's superior performance in clarification question generation and ambiguity resolution. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_15739 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ECLAIR: Enhanced Clarification for Interactive Responses Murzaku, John Liu, Zifan Tanjim, Md Mehrab Muppala, Vaishnavi Chen, Xiang Li, Yunyao Artificial Intelligence 68T50 I.2.7; H.5.2 We present ECLAIR (Enhanced CLArification for Interactive Responses), a novel unified and end-to-end framework for interactive disambiguation in enterprise AI assistants. ECLAIR generates clarification questions for ambiguous user queries and resolves ambiguity based on the user's response.We introduce a generalized architecture capable of integrating ambiguity information from multiple downstream agents, enhancing context-awareness in resolving ambiguities and allowing enterprise specific definition of agents. We further define agents within our system that provide domain-specific grounding information. We conduct experiments comparing ECLAIR to few-shot prompting techniques and demonstrate ECLAIR's superior performance in clarification question generation and ambiguity resolution. |
| title | ECLAIR: Enhanced Clarification for Interactive Responses |
| topic | Artificial Intelligence 68T50 I.2.7; H.5.2 |
| url | https://arxiv.org/abs/2503.15739 |