We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.IR

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Retrieval

Title: Adherence and Constancy in LIME-RS Explanations for Recommendation

Abstract: Explainable Recommendation has attracted a lot of attention due to a renewed interest in explainable artificial intelligence. In particular, post-hoc approaches have proved to be the most easily applicable ones to increasingly complex recommendation models, which are then treated as black-boxes. The most recent literature has shown that for post-hoc explanations based on local surrogate models, there are problems related to the robustness of the approach itself. This consideration becomes even more relevant in human-related tasks like recommendation. The explanation also has the arduous task of enhancing increasingly relevant aspects of user experience such as transparency or trustworthiness. This paper aims to show how the characteristics of a classical post-hoc model based on surrogates is strongly model-dependent and does not prove to be accountable for the explanations generated.
Comments: accepted at KaRS Workshop @RecSys 2021
Subjects: Information Retrieval (cs.IR)
Cite as: arXiv:2109.00818 [cs.IR]
  (or arXiv:2109.00818v3 [cs.IR] for this version)

Submission history

From: Claudio Pomo [view email]
[v1] Thu, 2 Sep 2021 09:58:09 GMT (233kb,D)
[v2] Sun, 5 Sep 2021 14:44:16 GMT (234kb,D)
[v3] Fri, 8 Oct 2021 08:56:05 GMT (230kb,D)

Link back to: arXiv, form interface, contact.