In a conversational context, a user converses with a system through a sequence of natural-language questions, i.e., utterances. Starting from a given subject, the conversation evolves through sequences of user utterances and system replies. The retrieval of documents relevant to an utterance is difficult due to informal use of natural language in speech and the complexity of understanding the semantic context coming from previous utterances. We adopt the 2019 TREC Conversational Assistant Track (CAsT) framework to experiment with a modular architecture performing in order: (i) automatic utterance understanding and rewriting, (ii) first-stage retrieval of candidate passages for the rewritten utterances, and (iii) neural re-ranking of candidate passages. By understanding the conversational context, we propose adaptive utterance rewriting strategies based on the current utterance and the dialogue evolution of the user with the system. A classifier identifies those utterances lacking context information as well as the dependencies on the previous utterances. Experimentally, we evaluate the proposed architecture in terms of traditional information retrieval metrics at small cutoffs. Results demonstrate the effectiveness of our techniques, achieving an improvement up to 0.6512 ($+201\text{\%}$) for P@1 and 0.4484 ($+214\text{\%}$) for nDCG@3 w.r.t. the CAsT baseline.

在对话上下文中，用户通过一系列自然语言问题（即话语）与系统进行对话。从给定的主题开始，对话通过用户话语和系统回复的序列演变。由于自然语言在语音中的非正式使用以及理解来自先前话语的语义上下文的复杂性，检索与话语相关的文档很困难。我们采用了 2019 年 TREC 会话助理轨道 ( CAsT) 框架来试验模块化架构，按顺序执行：(i) 自动话语理解和重写，(ii) 重写话语的候选段落的第一阶段检索，以及 (iii) 候选段落的神经重新排序。通过理解会话上下文，我们提出了基于当前话语和用户与系统对话演变的自适应话语重写策略。分类器识别那些缺乏上下文信息的话语以及对先前话语的依赖。在实验上，我们根据传统信息检索指标在小截止值方面评估所提出的架构。结果证明了我们技术的有效性，实现了高达 0.6512 ($+201\text{\%}$) 对于 P@1 和 0.4484 ($+214\text{\%}$) 用于 nDCG@3 与CAsT基线。