Improving Conversational Recommendation Systems via Counterfactual Data Simulation

TL;DR

This work tackles data scarcity in conversational recommender systems by introducing CFCRS, a counterfactual data simulation framework. It combines a knowledge-graph–aided user-preference representation with a flow language model to generate coherent conversation flows that are realized into dialogues, and then applies curriculum adversarial learning to augment data through controllable edits of user preferences. The approach demonstrates consistent improvements over strong baselines and other augmentation methods on ReDial and INSPIRED, with especially notable gains in low-data regimes, highlighting improved data efficiency. The proposed CFCRS framework offers a practical, model-agnostic path to scalable, high-quality CRSs and points toward future integration with larger language models to further enhance performance.

Abstract

Conversational recommender systems (CRSs) aim to provide recommendation services via natural language conversations. Although a number of approaches have been proposed for developing capable CRSs, they typically rely on sufficient training data for training. Since it is difficult to annotate recommendation-oriented dialogue datasets, existing CRS approaches often suffer from the issue of insufficient training due to the scarcity of training data. To address this issue, in this paper, we propose a CounterFactual data simulation approach for CRS, named CFCRS, to alleviate the issue of data scarcity in CRSs. Our approach is developed based on the framework of counterfactual data augmentation, which gradually incorporates the rewriting to the user preference from a real dialogue without interfering with the entire conversation flow. To develop our approach, we characterize user preference and organize the conversation flow by the entities involved in the dialogue, and design a multi-stage recommendation dialogue simulator based on a conversation flow language model. Under the guidance of the learned user preference and dialogue schema, the flow language model can produce reasonable, coherent conversation flows, which can be further realized into complete dialogues. Based on the simulator, we perform the intervention at the representations of the interacted entities of target users, and design an adversarial training method with a curriculum schedule that can gradually optimize the data augmentation strategy. Extensive experiments show that our approach can consistently boost the performance of several competitive CRSs, and outperform other data augmentation methods, especially when the training data is limited. Our code is publicly available at https://github.com/RUCAIBox/CFCRS.

Figures (4)

• Figure 1: The overview of our approach CFCRS. We first adopt curriculum counterfactual learning to augment the user preference at the representation level, and then use the flow language model guided by user and schema prompts to generate conversation flows, which are then realized into dialogues. The edit function and CRS model are optimized with adversarial training to improve both the quality of the augmented data and the recommendation performance.
• Figure 2: Performance comparison w.r.t. different amounts of training data on the ReDial dataset. We implement our framework on KBRD.
• Figure 3: Performance comparison w.r.t. different ratios of augmented examples on ReDial and Inspired dataset. We implement our approach on KBRD.
• Figure 4: Performance comparison w.r.t. different weights of the L2-norm loss $\lambda$ on ReDial and Inspired dataset. We implement our approach on KBRD.