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

Statistical spoken dialogue systems have the attractive property of beingable to be optimised from data via interactions with real users. However in thereinforcement learning paradigm the dialogue manager (agent) often requiressignificant time to explore the state-action space to learn to behave in adesirable manner. This is a critical issue when the system is trained on-linewith real users where learning costs are expensive. Reward shaping is onepromising technique for addressing these concerns. Here we examine threerecurrent neural network (RNN) approaches for providing reward shapinginformation in addition to the primary (task-orientated) environmentalfeedback. These RNNs are trained on returns from dialogues generated by asimulated user and attempt to diffuse the overall evaluation of the dialogueback down to the turn level to guide the agent towards good behaviour faster.In both simulated and real user scenarios these RNNs are shown to increasepolicy learning speed. Importantly, they do not require prior knowledge of theuser's goal.

Reward Shaping with Recurrent Neural Networks for Speeding up On-Line Policy Learning in Spoken Dialogue Systems