English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Conversational Recommender Systems (CRS) has become an emerging researchtopic seeking to perform recommendations through interactive conversations,which generally consist of generation and recommendation modules. Prior work onCRS tends to incorporate more external and domain-specific knowledge like itemreviews to enhance performance. Despite the fact that the collection andannotation of the external domain-specific information needs much human effortand degenerates the generalizability, too much extra knowledge introduces moredifficulty to balance among them. Therefore, we propose to fully discover andextract internal knowledge from the context. We capture both entity-level andcontextual-level representations to jointly model user preferences for therecommendation, where a time-aware attention is designed to emphasize therecently appeared items in entity-level representations. We further use thepre-trained BART to initialize the generation module to alleviate the datascarcity and enhance the context modeling. In addition to conductingexperiments on a popular dataset (ReDial), we also include a multi-domaindataset (OpenDialKG) to show the effectiveness of our model. Experiments onboth datasets show that our model achieves better performance on mostevaluation metrics with less external knowledge and generalizes well to otherdomains. Additional analyses on the recommendation and generation tasksdemonstrate the effectiveness of our model in different scenarios.

Improving Conversational Recommender System via Contextual and Time-Aware Modeling with Less Domain-Specific Knowledge