Introduction à l’étude du concept et de la signification lexicale de francophonie. Construction discursive d’un concept, activation d’un lien dénominatif, ou désignation d’un « objet social » ? (Olga Galatanu)

2016 ◽  
Keyword(s):  
Concept Activation

Chapter 15. Human-Centered Concept Explanations for Neural Networks

2021 ◽  
Author(s):  
Chih-Kuan Yeh ◽  
Been Kim ◽  
Pradeep Ravikumar
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Case Studies ◽  
Real World ◽  
Deep Neural Networks ◽  
Learning Models ◽  
Real World Applications ◽  
The Right ◽  
Concept Activation ◽  
Machine Learning Models

Understanding complex machine learning models such as deep neural networks with explanations is crucial in various applications. Many explanations stem from the model perspective, and may not necessarily effectively communicate why the model is making its predictions at the right level of abstraction. For example, providing importance weights to individual pixels in an image can only express which parts of that particular image is important to the model, but humans may prefer an explanation which explains the prediction by concept-based thinking. In this work, we review the emerging area of concept based explanations. We start by introducing concept explanations including the class of Concept Activation Vectors (CAV) which characterize concepts using vectors in appropriate spaces of neural activations, and discuss different properties of useful concepts, and approaches to measure the usefulness of concept vectors. We then discuss approaches to automatically extract concepts, and approaches to address some of their caveats. Finally, we discuss some case studies that showcase the utility of such concept-based explanations in synthetic settings and real world applications.

scholarly journals Concept Discovery for The Interpretation of Landscape Scenicness

2020 ◽  
Vol 2 (4) ◽  
pp. 397-413
Author(s):  
Pim Arendsen ◽  
Diego Marcos ◽  
Devis Tuia
Keyword(s):  
Vector Spaces ◽  
Large Set ◽  
Feature Representations ◽  
The United Kingdom ◽  
New Concepts ◽  
Manifold Alignment ◽  
Semantic Concepts ◽  
Visual Concepts ◽  
Alignment Technique ◽  
Concept Activation

In this paper, we study how to extract visual concepts to understand landscape scenicness. Using visual feature representations from a Convolutional Neural Network (CNN), we learn a number of Concept Activation Vectors (CAV) aligned with semantic concepts from ancillary datasets. These concepts represent objects, attributes or scene categories that describe outdoor images. We then use these CAVs to study their impact on the (crowdsourced) perception of beauty of landscapes in the United Kingdom. Finally, we deploy a technique to explore new concepts beyond those initially available in the ancillary dataset: Using a semi-supervised manifold alignment technique, we align the CNN image representation to a large set of word embeddings, therefore giving access to entire dictionaries of concepts. This allows us to obtain a list of new concept candidates to improve our understanding of the elements that contribute the most to the perception of scenicness. We do this without the need for any additional data by leveraging the commonalities in the visual and word vector spaces. Our results suggest that new and potentially useful concepts can be discovered by leveraging neighbourhood structures in the word vector spaces.

scholarly journals On Interpretability of Deep Learning based Skin Lesion Classifiers using Concept Activation Vectors

2020 ◽  
Author(s):  
Adriano Lucieri ◽  
Muhammad Naseer Bajwa ◽  
Stephan Alexander Braun ◽  
Muhammad Imran Malik ◽  
Andreas Dengel ◽  
...  
Keyword(s):  
Deep Learning ◽  
Skin Lesion ◽  
Concept Activation

Code-Switching Across Brainstorming Sessions: Implications for the Revised Hierarchical Model of Bilingual Language Processing

2003 ◽  
Vol 50 (3) ◽  
pp. 171-183 ◽  
Author(s):  
Kevin J. Blot ◽  
Michael A. Zárate ◽  
Paul B. Paulus
Keyword(s):  
Language Processing ◽  
Hierarchical Model ◽  
Word Form ◽  
Dominant Language ◽  
Following Instructions ◽  
Word Forms ◽  
Dual Activation ◽  
Concept Activation ◽  
L1 And L2 ◽  
Strong Concept

Abstract. The revised hierarchical model (RHM) of bilingual language processing posits independent word form representations for the dominant language (L1) and the nondominant language (L2), facilitated translation from L2 words to L1 words, access to common concepts for L1 and L2, and stronger activation of concepts for L1 than for L2. Spanish-English and English-Spanish bilinguals brainstormed for two sessions; half switched languages (L1-L2 or L2-L1) and half stayed in the same language (L1-L1 or L2-L2) across sessions. In both sessions, L1 brainstorming resulted in more efficient idea productivity than L2 brainstorming, supporting stronger concept activation for L1, consistent with the RHM. Switching languages from L2 to L1 resulted in the most efficient idea productivity in Session 2, suggesting that switching to L1 not only permits strong concept activation, but also the activation of concepts that are relatively different than those activated by L2, inconsistent with the RHM. Switching languages increased the proportion of Session 1 ideas repeated during Session 2, despite instructions not to repeat. This finding suggests that there is activation of concepts as well as word forms in same language brainstorming and that this dual activation aids in following instructions not to repeat, consistent with the RHM. It is suggested that the RHM be re-specified to accommodate the notion that L1 and L2 access relatively different concepts.

Follower Self-Concept Activation as the Process Underlying Leadership Effects

2006 ◽  
Author(s):  
Chu-Hsiang Chang ◽  
Rosalie J. Hall ◽  
Russell E. Johnson
Keyword(s):  
Self Concept ◽  
Concept Activation
Sign in / Sign up

Export Citation Format

Share Document