ImpReSS: Implicit Recommender System for Support Conversations

2025-06-17

AI Summary

Existing conversational recommendation systems (CRS) in customer service require explicit user intent (e.g., purchase signals) or interactive interventions to trigger recommendations, hindering seamless, implicit integration of recommendation into problem-solving dialogues. Method: This paper proposes the first fully implicit conversational recommendation framework—eliminating reliance on explicit purchase intent or user interaction—and instead automatically identifies Solution Product Categories (SPCs) solely from user queries via dialogue understanding. It leverages a domain-adapted large language model to jointly perform fine-grained SPC classification and ranking. Contribution/Results: The framework breaks the conventional CRS dependency on explicit intent modeling, enabling tight synergy between recommendation and problem resolution while supporting both service enhancement and business growth. Evaluated on three real-world troubleshooting tasks—general issue resolution, information security, and cybersecurity—the framework achieves MRR@1 of 0.72–0.85 and Recall@3 of 0.67–0.89.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Conversational AI/Dialog Systems
• Machine Learning: Learning Preferences or Rankings

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Assisted, interactive, and conversational search
• Responsible Web: Consent frameworks and practices on the web

Abstract

Following recent advancements in large language models (LLMs), LLM-based chatbots have transformed customer support by automating interactions and providing consistent, scalable service. While LLM-based conversational recommender systems (CRSs) have attracted attention for their ability to enhance the quality of recommendations, limited research has addressed the implicit integration of recommendations within customer support interactions. In this work, we introduce ImpReSS, an implicit recommender system designed for customer support conversations. ImpReSS operates alongside existing support chatbots, where users report issues and chatbots provide solutions. Based on a customer support conversation, ImpReSS identifies opportunities to recommend relevant solution product categories (SPCs) that help resolve the issue or prevent its recurrence -- thereby also supporting business growth. Unlike traditional CRSs, ImpReSS functions entirely implicitly and does not rely on any assumption of a user's purchasing intent. Our empirical evaluation of ImpReSS's ability to recommend relevant SPCs that can help address issues raised in support conversations shows promising results, including an MRR@1 (and recall@3) of 0.72 (0.89) for general problem solving, 0.82 (0.83) for information security support, and 0.85 (0.67) for cybersecurity troubleshooting. To support future research, our data and code will be shared upon request.

Problem

• Implicitly integrate recommendations in support conversations
• Recommend solution products without user purchase intent
• Enhance issue resolution and prevent recurrence via SPCs

Innovation

• Implicit integration of recommendations in support chats
• LLM-based chatbot for automated customer support
• Recommends solution product categories without purchase intent