Recently, Logic Explained Networks (LENs) have been proposed as explainable-by-design neural models providing logic explanations for their predictions. However, these models have only been applied to vision and tabular data, and they mostly favour the generation of global explanations, while local ones tend to be noisy and verbose. For these reasons, we propose LENp, improving local explanations by perturbing input words, and we test it on text classification. Our results show that (i) LENp provides better local explanations than LIME in terms of sensitivity and faithfulness, and (ii) its logic explanations are more useful and user-friendly than the feature scoring provided by LIME as attested by a human survey.
Extending Logic Explained Networks to Text Classification / Jain, Rishabh; Ciravegna, Gabriele; Barbiero, Pietro; Giannini, Francesco; Buffelli, Davide; Lio, Pietro. - ELETTRONICO. - (2022), pp. 8838-8857. (Intervento presentato al convegno 2022 Conference on Empirical Methods in Natural Language Processing tenutosi a Abu Dhabi (United Arab Emirates) nel 7 December 2022 -11 December 2022).
Extending Logic Explained Networks to Text Classification
Ciravegna, Gabriele;
2022
Abstract
Recently, Logic Explained Networks (LENs) have been proposed as explainable-by-design neural models providing logic explanations for their predictions. However, these models have only been applied to vision and tabular data, and they mostly favour the generation of global explanations, while local ones tend to be noisy and verbose. For these reasons, we propose LENp, improving local explanations by perturbing input words, and we test it on text classification. Our results show that (i) LENp provides better local explanations than LIME in terms of sensitivity and faithfulness, and (ii) its logic explanations are more useful and user-friendly than the feature scoring provided by LIME as attested by a human survey.File | Dimensione | Formato | |
---|---|---|---|
2022.emnlp-main.604.pdf
accesso aperto
Tipologia:
2a Post-print versione editoriale / Version of Record
Licenza:
Creative commons
Dimensione
779.18 kB
Formato
Adobe PDF
|
779.18 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/2980677