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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The ingrained principles of fairness in a dialogue system's decision-makingprocess and generated responses are crucial for user engagement, satisfaction,and task achievement. Absence of equitable and inclusive principles can hinderthe formation of common ground, which in turn negatively impacts the overallperformance of the system. For example, misusing pronouns in a user interactionmay cause ambiguity about the intended subject. Yet, there is no comprehensivestudy of equitable text generation in dialogue. Aptly, in this work, we usetheories of computational learning to study this problem. We provide formaldefinitions of equity in text generation, and further, prove formal connectionsbetween learning human-likeness and learning equity: algorithms for improvingequity ultimately reduce to algorithms for improving human-likeness (onaugmented data). With this insight, we also formulate reasonable conditionsunder which text generation algorithms can learn to generate equitable textwithout any modifications to the biased training data on which they learn. Toexemplify our theory in practice, we look at a group of algorithms for theGuessWhat?! visual dialogue game and, using this example, test our theoryempirically. Our theory accurately predicts relative-performance of multiplealgorithms in generating equitable text as measured by both human and automatedevaluation.

Learning to Generate Equitable Text in Dialogue from Biased Training Data