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

Despite recent advancements in language models (LMs), their application todialogue management (DM) problems and ability to carry on rich conversationsremain a challenge. We use reinforcement learning (RL) to develop a dialogueagent that avoids being short-sighted (outputting generic utterances) andmaximizes overall user satisfaction. Most existing RL approaches to DM trainthe agent at the word-level, and thus, have to deal with a combinatoriallycomplex action space even for a medium-size vocabulary. As a result, theystruggle to produce a successful and engaging dialogue even if they arewarm-started with a pre-trained LM. To address this issue, we develop aRL-based DM using a novel mixture of expert language model (MoE-LM) thatconsists of (i) a LM capable of learning diverse semantics for conversationhistories, (ii) a number of {\em specialized} LMs (or experts) capable ofgenerating utterances corresponding to a particular attribute or personality,and (iii) a RL-based DM that performs dialogue planning with the utterancesgenerated by the experts. Our MoE approach provides greater flexibility togenerate sensible utterances with different intents and allows RL to focus onconversational-level DM. We compare it with SOTA baselines on open-domaindialogues and demonstrate its effectiveness both in terms of the diversity andsensibility of the generated utterances and the overall DM performance.

A Mixture-of-Expert Approach to RL-based Dialogue Management