Improving Items and Contexts Understanding with Descriptive Graph for Conversational Recommendation

TL;DR

KLEVER is a new conversational recommender system that jointly models items and contextual words in a shared semantic space using a descriptive graph, improving recommendation quality especially with limited user input.

Contribution

This work introduces KLEVER, a framework that constructs an item descriptive graph to unify item and word representations, addressing semantic misalignment in CRS.

Findings

KLEVER outperforms existing CRS methods on benchmark datasets.

It maintains high recommendation accuracy with limited user input.

Joint modeling of items and words enhances dialog generation quality.

Abstract

State-of-the-art methods on conversational recommender systems (CRS) leverage external knowledge to enhance both items' and contextual words' representations to achieve high quality recommendations and responses generation. However, the representations of the items and words are usually modeled in two separated semantic spaces, which leads to misalignment issue between them. Consequently, this will cause the CRS to only achieve a sub-optimal ranking performance, especially when there is a lack of sufficient information from the user's input. To address limitations of previous works, we propose a new CRS framework KLEVER, which jointly models items and their associated contextual words in the same semantic space. Particularly, we construct an item descriptive graph from the rich items' textual features, such as item description and categories. Based on the constructed descriptive graph, KLEVER jointly learns the embeddings of the words and items, towards enhancing both recommender and dialog generation modules. Extensive experiments on benchmarking CRS dataset demonstrate that KLEVER achieves superior performance, especially when the information from the users' responses is lacking.