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/

Zendy Plus

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

Zendy Tools

Zendy Open

Nowadays, the use of machine learning models is becoming a utility in manyapplications. Companies deliver pre-trained models encapsulated as applicationprogramming interfaces (APIs) that developers combine with third partycomponents and their own models and data to create complex data products tosolve specific problems. The complexity of such products and the lack ofcontrol and knowledge of the internals of each component used cause unavoidableeffects, such as lack of transparency, difficulty in auditability, andemergence of potential uncontrolled risks. They are effectively black-boxes.Accountability of such solutions is a challenge for the auditors and themachine learning community. In this work, we propose a wrapper that given ablack-box model enriches its output prediction with a measure of uncertainty.By using this wrapper, we make the black-box auditable for the accuracy risk(risk derived from low quality or uncertain decisions) and at the same time weprovide an actionable mechanism to mitigate that risk in the form of decisionrejection; we can choose not to issue a prediction when the risk or uncertaintyin that decision is significant. Based on the resulting uncertainty measure, weadvocate for a rejection system that selects the more confident predictions,discarding those more uncertain, leading to an improvement in the trustabilityof the resulting system. We showcase the proposed technique and methodology ina practical scenario where a simulated sentiment analysis API based on naturallanguage processing is applied to different domains. Results demonstrate theeffectiveness of the uncertainty computed by the wrapper and its highcorrelation to bad quality predictions and misclassifications.

Dirichlet uncertainty wrappers for actionable algorithm accuracy accountability and auditability