scholarly journals Explanation-Based Human Debugging of NLP Models: A Survey

2021 ◽  
Vol 9 ◽  
pp. 1508-1528
Author(s):  
Piyawat Lertvittayakumjorn ◽  
Francesca Toni
Keyword(s):  
Learning Model ◽  
Training Data ◽  
Three Dimensions ◽  
Future Research ◽  
Experimental Setting ◽  
Open Problems ◽  
Research Directions ◽  
Machine Learning Model ◽  
Future Research Directions ◽  
Deep Learning Model

Abstract Debugging a machine learning model is hard since the bug usually involves the training data and the learning process. This becomes even harder for an opaque deep learning model if we have no clue about how the model actually works. In this survey, we review papers that exploit explanations to enable humans to give feedback and debug NLP models. We call this problem explanation-based human debugging (EBHD). In particular, we categorize and discuss existing work along three dimensions of EBHD (the bug context, the workflow, and the experimental setting), compile findings on how EBHD components affect the feedback providers, and highlight open problems that could be future research directions.

Metric Methods in Data Mining

2008 ◽  
pp. 849-879
Author(s):  
Dan A. Simovici
Keyword(s):  
Data Mining ◽  
Metric Space ◽  
Training Data ◽  
Future Research ◽  
Open Problems ◽  
Research Directions ◽  
Data Mining Techniques ◽  
Future Research Directions ◽  
Major Data ◽  
Geometric Study

This chapter presents data mining techniques that make use of metrics defined on the set of partitions of finite sets. Partitions are naturally associated with object attributes and major data mining problem such as classification, clustering, and data preparation benefit from an algebraic and geometric study of the metric space of partitions. The metrics we find most useful are derived from a generalization of the entropic metric. We discuss techniques that produce smaller classifiers, allow incremental clustering of categorical data and help user to better prepare training data for constructing classifiers. Finally, we discuss open problems and future research directions.

Metric Methods in Data Mining

2011 ◽  
pp. 1-31 ◽  
Author(s):  
Dan A. Simovici
Keyword(s):  
Data Mining ◽  
Metric Space ◽  
Training Data ◽  
Future Research ◽  
Open Problems ◽  
Research Directions ◽  
Data Mining Techniques ◽  
Future Research Directions ◽  
Major Data ◽  
Geometric Study

This chapter presents data mining techniques that make use of metrics defined on the set of partitions of finite sets. Partitions are naturally associated with object attributes and major data mining problem such as classification, clustering, and data preparation benefit from an algebraic and geometric study of the metric space of partitions. The metrics we find most useful are derived from a generalization of the entropic metric. We discuss techniques that produce smaller classifiers, allow incremental clustering of categorical data and help user to better prepare training data for constructing classifiers. Finally, we discuss open problems and future research directions.

MetaLearning with Graph Neural Networks

2021 ◽  
Vol 23 (2) ◽  
pp. 13-22
Author(s):  
Debmalya Mandal ◽  
Sourav Medya ◽  
Brian Uzzi ◽  
Charu Aggarwal
Keyword(s):  
Neural Networks ◽  
Deep Neural Networks ◽  
Future Research ◽  
Open Problems ◽  
Research Directions ◽  
Graph Problems ◽  
Meta Learning ◽  
Comprehensive Survey ◽  
Future Research Directions ◽  
Graph Neural Networks

Graph Neural Networks (GNNs), a generalization of deep neural networks on graph data have been widely used in various domains, ranging from drug discovery to recommender systems. However, GNNs on such applications are limited when there are few available samples. Meta-learning has been an important framework to address the lack of samples in machine learning, and in recent years, researchers have started to apply meta-learning to GNNs. In this work, we provide a comprehensive survey of different metalearning approaches involving GNNs on various graph problems showing the power of using these two approaches together. We categorize the literature based on proposed architectures, shared representations, and applications. Finally, we discuss several exciting future research directions and open problems.

scholarly journals Android Malware Family Classification and Analysis: Current Status and Future Directions

Electronics ◽  
2020 ◽  
Vol 9 (6) ◽  
pp. 942 ◽  
Author(s):  
Fahad Alswaina ◽  
Khaled Elleithy
Keyword(s):  
Three Dimensions ◽  
Current Status ◽  
Future Research ◽  
Future Directions ◽  
Research Directions ◽  
Android Malware ◽  
Literature Reviews ◽  
Comprehensive Survey ◽  
Future Research Directions ◽  
Major Attention

Android receives major attention from security practitioners and researchers due to the influx number of malicious applications. For the past twelve years, Android malicious applications have been grouped into families. In the research community, detecting new malware families is a challenge. As we investigate, most of the literature reviews focus on surveying malware detection. Characterizing the malware families can improve the detection process and understand the malware patterns. For this reason, we conduct a comprehensive survey on the state-of-the-art Android malware familial detection, identification, and categorization techniques. We categorize the literature based on three dimensions: type of analysis, features, and methodologies and techniques. Furthermore, we report the datasets that are commonly used. Finally, we highlight the limitations that we identify in the literature, challenges, and future research directions regarding the Android malware family.

What Motivates People to Customize Apparel Online?

2015 ◽  
pp. 194-210
Author(s):  
Hira Cho
Keyword(s):  
Three Dimensions ◽  
Female College Students ◽  
Future Research ◽  
Research Directions ◽  
Five Dimensions ◽  
Time Consumption ◽  
Apparel Shopping ◽  
Future Research Directions ◽  
Value Creation Process ◽  
The U.S

The adoption of customizing features is expected to provide a strategic advantage to e-retailers that want to move forward in a competitive environment. The goal of this chapter is to identify a variety of aspects of consumer perceptions on e-customization for apparel shopping and to understand what can motivate the willingness of consumers to participate in the value creation process. A survey was conducted using developed customization websites for the ordering of a pair of jeans. Data were collected from 213 female college students in the U.S. Their statements after experiencing the customization process were analyzed and categorized into three dimensions of benefits (usefulness, convenience, and fun/enjoyment) and five dimensions of costs (risk, limitation, self-assurance, time consumption, and unappealing) of e-customization. Based on the findings, two discussion topics were drawn: why people are willing or unwilling to customize apparel online. Insights are generated and future research directions are discussed.

Online Behaviour as Predictive of Professional Online Work Readiness Among Mass Communication Students

2020 ◽  
Vol 12 (2) ◽  
pp. 48-61
Author(s):  
Joseph Njuguna ◽  
Margaret Jjuuko
Keyword(s):  
Mass Communication ◽  
Three Dimensions ◽  
Future Research ◽  
Audience Engagement ◽  
Research Directions ◽  
Significant Difference ◽  
Online Behaviour ◽  
Future Research Directions ◽  
Online Skills ◽  
The Relationship

This study explores the relationship between students' online behaviour and their perceived readiness for professional online work. A sample of mass communication students (n=143) from five Rwandan universities completed a questionnaire. Analysis showed that the three dimensions of online behaviour correlated positively with the students' perceived readiness for professional online work. While the respondents' audience engagement levels had the highest correlation with online skills readiness, the frequency of online accounts usage had the weakest correlation. The multiple regression results revealed that levels of engagement and ownership of online accounts were significant predictors of the students' perceived professional online skills readiness. Furthermore, there was a statistically significant difference in perceived professional online skills readiness between students who rated themselves as highly prepared and those who expressed low levels of preparedness for professional online skills. Implications and future research directions are discussed based on the findings.

scholarly journals Organizational Trust: A Gateway to Psychological Empowerment

2016 ◽  
Vol 9 (1) ◽  
pp. 52 ◽  
Author(s):  
Sanad A. Alajmi
Keyword(s):  
Psychological Empowerment ◽  
Organizational Trust ◽  
Industrial Area ◽  
Two Dimensions ◽  
Three Dimensions ◽  
Future Research ◽  
Research Directions ◽  
Industrial Companies ◽  
Future Research Directions ◽  
The Relationship

<p>This study investigates the relationship between psychological empowerment of employees and organizational trust within Kuwaiti industrial companies. It focuses on two dimensions of organizational trust; namely, trust in supervisors and trust in the organization. A total of 450 questionnaires were submitted to industrial companies in the Subhan Industrial Area, of which350 were completed. The results indicate that a significant positive correlation exists between psychological empowerment of the employees of these companies and organizational trust. The findings indicate that trust in the supervisor and in the organizationexplains21.8% and 13.1%, respectively, of the variation in psychological empowerment. Trust in the supervisor correlates significantly and positively with all dimensions of psychological empowerment whereas trust in the organization correlates significantly and positively with only three dimensions of psychological empowerment; namely, meaning, competence, and self-determination. The study concludes by explaining the limitations involved and suggests future research directions to enhance psychological empowerment and trust in industrial companies in Kuwait.</p>

Interpreting Deep Learning-based Vulnerability Detector Predictions Based on Heuristic Searching

2021 ◽  
Vol 30 (2) ◽  
pp. 1-31
Author(s):  
Deqing Zou ◽  
Yawei Zhu ◽  
Shouhuai Xu ◽  
Zhen Li ◽  
Hai Jin ◽  
...  
Keyword(s):  
Deep Learning ◽  
Learning Model ◽  
Future Research ◽  
Learning Models ◽  
Vulnerability Detection ◽  
Open Problems ◽  
Domain Experts ◽  
Software Vulnerabilities ◽  
Significant Step ◽  
Deep Learning Model

Detecting software vulnerabilities is an important problem and a recent development in tackling the problem is the use of deep learning models to detect software vulnerabilities. While effective, it is hard to explain why a deep learning model predicts a piece of code as vulnerable or not because of the black-box nature of deep learning models. Indeed, the interpretability of deep learning models is a daunting open problem. In this article, we make a significant step toward tackling the interpretability of deep learning model in vulnerability detection. Specifically, we introduce a high-fidelity explanation framework, which aims to identify a small number of tokens that make significant contributions to a detector’s prediction with respect to an example. Systematic experiments show that the framework indeed has a higher fidelity than existing methods, especially when features are not independent of each other (which often occurs in the real world). In particular, the framework can produce some vulnerability rules that can be understood by domain experts for accepting a detector’s outputs (i.e., true positives) or rejecting a detector’s outputs (i.e., false-positives and false-negatives). We also discuss limitations of the present study, which indicate interesting open problems for future research.

scholarly journals Sensor-Based Control for Collaborative Robots: Fundamentals, Challenges, and Opportunities

2021 ◽  
Vol 14 ◽  
Author(s):  
Andrea Cherubini ◽  
David Navarro-Alarcon
Keyword(s):  
Integration Method ◽  
Control Level ◽  
Direct Interaction ◽  
Future Research ◽  
Open Problems ◽  
Research Directions ◽  
Heterogeneous Sensors ◽  
Challenges And Opportunities ◽  
Potential Applications ◽  
Future Research Directions

The objective of this paper is to present a systematic review of existing sensor-based control methodologies for applications that involve direct interaction between humans and robots, in the form of either physical collaboration or safe coexistence. To this end, we first introduce the basic formulation of the sensor-servo problem, and then, present its most common approaches: vision-based, touch-based, audio-based, and distance-based control. Afterwards, we discuss and formalize the methods that integrate heterogeneous sensors at the control level. The surveyed body of literature is classified according to various factors such as: sensor type, sensor integration method, and application domain. Finally, we discuss open problems, potential applications, and future research directions.

Sign in / Sign up

Export Citation Format

Share Document