We present a methodology to systematically test conversational recommender systems with regards to conversational breakdowns. It involves examining conversations generated between the system and simulated users for a set of pre-defined breakdown types, extracting responsible conversational paths, and characterizing them in terms of the underlying dialogue intents. User simulation offers the advantages of simplicity, cost-effectiveness, and time efficiency for obtaining conversations where potential breakdowns can be identified. The proposed methodology can be used as diagnostic tool as well as a development tool to improve conversational recommendation systems. We apply our methodology in a case study with an existing conversational recommender system and user simulator, demonstrating that with just a few iterations, we can make the system more robust to conversational breakdowns.
While recommender systems with multi-modal item representations (image, audio, and text), have been widely explored, learning recommendations from multi-modal user interactions (e.g., clicks and speech) remains an open problem. We study the case of multi-modal user interactions in a setting where users engage with a service provider through multiple channels (website and call center). In such cases, incomplete modalities naturally occur, since not all users interact through all the available channels. To address these challenges, we publish a real-world dataset that allows progress in this under-researched area. We further present and benchmark various methods for leveraging multi-modal user interactions for item recommendations, and propose a novel approach that specifically deals with missing modalities by mapping user interactions to a common feature space. Our analysis reveals important interactions between the different modalities and that a frequently occurring modality can enhance learning from a less frequent one.
The increasing reliance on digital information necessitates advancements in conversational search systems, particularly in terms of information transparency. While prior research in conversational information-seeking has concentrated on improving retrieval techniques, the challenge remains in generating responses useful from a user perspective. This study explores different methods of explaining the responses, hypothesizing that transparency about the source of the information, system confidence, and limitations can enhance users' ability to objectively assess the response. By exploring transparency across explanation type, quality, and presentation mode, this research aims to bridge the gap between system-generated responses and responses verifiable by the user. We design a user study to answer questions concerning the impact of (1) the quality of explanations enhancing the response on its usefulness and (2) ways of presenting explanations to users. The analysis of the collected data reveals lower user ratings for noisy explanations, although these scores seem insensitive to the quality of the response. Inconclusive results on the explanations presentation format suggest that it may not be a critical factor in this setting.
Conversational information-seeking (CIS) is an emerging paradigm for knowledge acquisition and exploratory search. Traditional web search interfaces enable easy exploration of entities, but this is limited in conversational settings due to the limited-bandwidth interface. This paper explore ways to rewrite answers in CIS, so that users can understand them without having to resort to external services or sources. Specifically, we focus on salient entities -- entities that are central to understanding the answer. As our first contribution, we create a dataset of conversations annotated with entities for saliency. Our analysis of the collected data reveals that the majority of answers contain salient entities. As our second contribution, we propose two answer rewriting strategies aimed at improving the overall user experience in CIS. One approach expands answers with inline definitions of salient entities, making the answer self-contained. The other approach complements answers with follow-up questions, offering users the possibility to learn more about specific entities. Results of a crowdsourcing-based study indicate that rewritten answers are clearly preferred over the original ones. We also find that inline definitions tend to be favored over follow-up questions, but this choice is highly subjective, thereby providing a promising future direction for personalization.
While interest in conversational recommender systems has been on the rise, operational systems suitable for serving as research platforms for comprehensive studies are currently lacking. This paper introduces an enhanced version of the IAI MovieBot conversational movie recommender system, aiming to evolve it into a robust and adaptable platform for conducting user-facing experiments. The key highlights of this enhancement include the addition of trainable neural components for natural language understanding and dialogue policy, transparent and explainable modeling of user preferences, along with improvements in the user interface and research infrastructure.
Personal knowledge graphs (PKGs) offer individuals a way to store and consolidate their fragmented personal data in a central place, improving service personalization while maintaining full user control. Despite their potential, practical PKG implementations with user-friendly interfaces remain scarce. This work addresses this gap by proposing a complete solution to represent, manage, and interface with PKGs. Our approach includes (1) a user-facing PKG Client, enabling end-users to administer their personal data easily via natural language statements, and (2) a service-oriented PKG API. To tackle the complexity of representing these statements within a PKG, we present an RDF-based PKG vocabulary that supports this, along with properties for access rights and provenance.
While the body of research directed towards constructing and generating clarifying questions in mixed-initiative conversational search systems is vast, research aimed at processing and comprehending users' answers to such questions is scarce. To this end, we present a simple yet effective method for processing answers to clarifying questions, moving away from previous work that simply appends answers to the original query and thus potentially degrades retrieval performance. Specifically, we propose a classifier for assessing usefulness of the prompted clarifying question and an answer given by the user. Useful questions or answers are further appended to the conversation history and passed to a transformer-based query rewriting module. Results demonstrate significant improvements over strong non-mixed-initiative baselines. Furthermore, the proposed approach mitigates the performance drops when non useful questions and answers are utilized.
Generative AI models face the challenge of hallucinations that can undermine users' trust in such systems. We approach the problem of conversational information seeking as a two-step process, where relevant passages in a corpus are identified first and then summarized into a final system response. This way we can automatically assess if the answer to the user's question is present in the corpus. Specifically, our proposed method employs a sentence-level classifier to detect if the answer is present, then aggregates these predictions on the passage level, and eventually across the top-ranked passages to arrive at a final answerability estimate. For training and evaluation, we develop a dataset based on the TREC CAsT benchmark that includes answerability labels on the sentence, passage, and ranking levels. We demonstrate that our proposed method represents a strong baseline and outperforms a state-of-the-art LLM on the answerability prediction task.
Research on conversational search has so far mostly focused on query rewriting and multi-stage passage retrieval. However, synthesizing the top retrieved passages into a complete, relevant, and concise response is still an open challenge. Having snippet-level annotations of relevant passages would enable both (1) the training of response generation models that are able to ground answers in actual statements and (2) the automatic evaluation of the generated responses in terms of completeness. In this paper, we address the problem of collecting high-quality snippet-level answer annotations for two of the TREC Conversational Assistance track datasets. To ensure quality, we first perform a preliminary annotation study, employing different task designs, crowdsourcing platforms, and workers with different qualifications. Based on the outcomes of this study, we refine our annotation protocol before proceeding with the full-scale data collection. Overall, we gather annotations for 1.8k question-paragraph pairs, each annotated by three independent crowd workers. The process of collecting data at this magnitude also led to multiple insights about the problem that can inform the design of future response-generation methods. This is an extended version of the article published with the same title in the Proceedings of CIKM'23.
Traditional recommender systems leverage users' item preference history to recommend novel content that users may like. However, modern dialog interfaces that allow users to express language-based preferences offer a fundamentally different modality for preference input. Inspired by recent successes of prompting paradigms for large language models (LLMs), we study their use for making recommendations from both item-based and language-based preferences in comparison to state-of-the-art item-based collaborative filtering (CF) methods. To support this investigation, we collect a new dataset consisting of both item-based and language-based preferences elicited from users along with their ratings on a variety of (biased) recommended items and (unbiased) random items. Among numerous experimental results, we find that LLMs provide competitive recommendation performance for pure language-based preferences (no item preferences) in the near cold-start case in comparison to item-based CF methods, despite having no supervised training for this specific task (zero-shot) or only a few labels (few-shot). This is particularly promising as language-based preference representations are more explainable and scrutable than item-based or vector-based representations.