A Clarifying Question Selection System from NTES_ALONG in Convai3 Challenge

TL;DR

This paper introduces a system for selecting clarification questions in conversational AI, combining response understanding, candidate recall, and ranking models, achieving top scores in relevance metrics.

Contribution

We propose an ensemble of fine-tuned RoBERTa and ELECTRA models for clarification question selection in conversational AI, demonstrating state-of-the-art performance.

Findings

Ensemble model outperforms individual models in relevance tasks.

Achieves top recall@[20,30] metrics in question relevance.

Attains highest scores in multi-turn conversation evaluation.

Abstract

This paper presents the participation of NetEase Game AI Lab team for the ClariQ challenge at Search-oriented Conversational AI (SCAI) EMNLP workshop in 2020. The challenge asks for a complete conversational information retrieval system that can understanding and generating clarification questions. We propose a clarifying question selection system which consists of response understanding, candidate question recalling and clarifying question ranking. We fine-tune a RoBERTa model to understand user's responses and use an enhanced BM25 model to recall the candidate questions. In clarifying question ranking stage, we reconstruct the training dataset and propose two models based on ELECTRA. Finally we ensemble the models by summing up their output probabilities and choose the question with the highest probability as the clarification question. Experiments show that our ensemble ranking model outperforms in the document relevance task and achieves the best recall@[20,30] metrics in question relevance task. And in multi-turn conversation evaluation in stage2, our system achieve the top score of all document relevance metrics.