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 the high accuracy offered by state-of-the-art deep natural-languagemodels (e.g. LSTM, BERT), their application in real-life settings is stillwidely limited, as they behave like a black-box to the end-user. Hence,explainability is rapidly becoming a fundamental requirement offuture-generation data-driven systems based on deep-learning approaches.Several attempts to fulfill the existing gap between accuracy andinterpretability have been done. However, robust and specialized xAI(Explainable Artificial Intelligence) solutions tailored to deepnatural-language models are still missing. We propose a new framework, namedT-EBAnO, which provides innovative prediction-local and class-basedmodel-global explanation strategies tailored to black-box deep natural-languagemodels. Given a deep NLP model and the textual input data, T-EBAnO provides anobjective, human-readable, domain-specific assessment of the reasons behind theautomatic decision-making process. Specifically, the framework extracts sets ofinterpretable features mining the inner knowledge of the model. Then, itquantifies the influence of each feature during the prediction process byexploiting the novel normalized Perturbation Influence Relation index at thelocal level and the novel Global Absolute Influence and Global RelativeInfluence indexes at the global level. The effectiveness and the quality of thelocal and global explanations obtained with T-EBAnO are proved on (i) asentiment analysis task performed by a fine-tuned BERT model, and (ii) a toxiccomment classification task performed by an LSTM model.

Explaining the Deep Natural Language Processing by Mining Textual Interpretable Features