Passage Query Methods for Retrieval and Reranking in Conversational Agents
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909522249908224 |
|---|---|
| author | De Lima, Victor Yang, Grace Hui |
| author_facet | De Lima, Victor Yang, Grace Hui |
| contents | This paper presents our approach to the TREC Interactive Knowledge Assistance Track (iKAT), which focuses on improving conversational information-seeking (CIS) systems. While recent advancements in CIS have improved conversational agents' ability to assist users, significant challenges remain in understanding context and retrieving relevant documents across domains and dialogue turns. To address these issues, we extend the Generate-Retrieve-Generate pipeline by developing passage queries (PQs) that align with the target document's expected format to improve query-document matching during retrieval. We propose two variations of this approach: Weighted Reranking and Short and Long Passages. Each method leverages a Meta Llama model for context understanding and generating queries and responses. Passage ranking evaluation results show that the Short and Long Passages approach outperformed the organizers' baselines, performed best among Llama-based systems in the track, and achieved results comparable to GPT-4-based systems. These results indicate that the method effectively balances efficiency and performance. Findings suggest that PQs improve semantic alignment with target documents and demonstrate their potential to improve multi-turn dialogue systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_00238 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Passage Query Methods for Retrieval and Reranking in Conversational Agents De Lima, Victor Yang, Grace Hui Information Retrieval This paper presents our approach to the TREC Interactive Knowledge Assistance Track (iKAT), which focuses on improving conversational information-seeking (CIS) systems. While recent advancements in CIS have improved conversational agents' ability to assist users, significant challenges remain in understanding context and retrieving relevant documents across domains and dialogue turns. To address these issues, we extend the Generate-Retrieve-Generate pipeline by developing passage queries (PQs) that align with the target document's expected format to improve query-document matching during retrieval. We propose two variations of this approach: Weighted Reranking and Short and Long Passages. Each method leverages a Meta Llama model for context understanding and generating queries and responses. Passage ranking evaluation results show that the Short and Long Passages approach outperformed the organizers' baselines, performed best among Llama-based systems in the track, and achieved results comparable to GPT-4-based systems. These results indicate that the method effectively balances efficiency and performance. Findings suggest that PQs improve semantic alignment with target documents and demonstrate their potential to improve multi-turn dialogue systems. |
| title | Passage Query Methods for Retrieval and Reranking in Conversational Agents |
| topic | Information Retrieval |
| url | https://arxiv.org/abs/2503.00238 |