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

Download:

Current browse context:

cs.CL

Change to browse by:

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 > Computation and Language

Title: Interpreting convolutional networks trained on textual data

Abstract: There have been many advances in the artificial intelligence field due to the emergence of deep learning. In almost all sub-fields, artificial neural networks have reached or exceeded human-level performance. However, most of the models are not interpretable. As a result, it is hard to trust their decisions, especially in life and death scenarios. In recent years, there has been a movement toward creating explainable artificial intelligence, but most work to date has concentrated on image processing models, as it is easier for humans to perceive visual patterns. There has been little work in other fields like natural language processing. In this paper, we train a convolutional model on textual data and analyze the global logic of the model by studying its filter values. In the end, we find the most important words in our corpus to our models logic and remove the rest (95%). New models trained on just the 5% most important words can achieve the same performance as the original model while reducing training time by more than half. Approaches such as this will help us to understand NLP models, explain their decisions according to their word choices, and improve them by finding blind spots and biases.
Comments: 9 pages, 6 figures, 5 tables
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
Cite as: arXiv:2010.13585 [cs.CL]
  (or arXiv:2010.13585v1 [cs.CL] for this version)

Submission history

From: Reza Marzban [view email]
[v1] Tue, 20 Oct 2020 20:12:05 GMT (9107kb,D)

Link back to: arXiv, form interface, contact.