Improving the Reliability of Deep Neural Networks in NLP: A Review

Alshemali, Basemah; Kalita, Jugal

doi:10.1016/j.knosys.2019.105210

Cited by 132 publications

(62 citation statements)

References 73 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep neural networks have been successfully employed for different types of machine-learning tasks, such as NLP-based methods utilizing sentiment aspects for deep classification [21]- [26]. Deep neural networks are able to model high-level abstractions and to decrease the dimensions by utilizing multiple processing layers based on complex structures or to be combined with non-linear transformations.…”

Section: Deep Learning and Covid-19-sentiment Classificationmentioning

confidence: 99%

Deep Sentiment Classification and Topic Discovery on Novel Coronavirus or COVID-19 Online Discussions: NLP Using LSTM Recurrent Neural Network Approach

Jelodar

Wang

Orji

et al. 2020

IEEE J. Biomed. Health Inform.

303

166

View full text Add to dashboard Cite

Internet forums and public social media, such as online healthcare forums, provide a convenient channel for users (people/patients) concerned about health issues to discuss and share information with each other. In late December 2019, an outbreak of a novel coronavirus (infection from which results in the disease named COVID-19) was reported, and, due to the rapid spread of the virus in other parts of the world, the World Health Organization declared a state of emergency. In this paper, we used automated extraction of COVID-19-related discussions from social media and a natural language process (NLP) method based on topic modeling to uncover various issues related to COVID-19 from public opinions. Moreover, we also investigate how to use LSTM recurrent neural network for sentiment classification of COVID-19 comments. Our findings shed light on the importance of using public opinions and suitable computational techniques to understand issues surrounding COVID-19 and to guide related decision-making. In addition, experiments demonstrated that the research model achieved an accuracy of 81.15%-a higher accuracy than that of several other well-known machine-learning algorithms for COVID-19-Sentiment Classification.

show abstract

Section: Deep Learning and Covid-19-sentiment Classificationmentioning

confidence: 99%

Deep Sentiment Classification and Topic Discovery on Novel Coronavirus or COVID-19 Online Discussions: NLP Using LSTM Recurrent Neural Network Approach

Jelodar

Wang

Orji

et al. 2020

IEEE J. Biomed. Health Inform.

303

166

View full text Add to dashboard Cite

show abstract

“…Explored in the mid 1950s Artificial Intelligence was aimed to build machines which imitate human intelligence and have understanding of human behavior (Ertel, 2018;Garnham, 2017). Over the time, Artficial Neural Networks have gained a considerable popularity for solving Natural language processing problems (Alshemali & Kalita, 2020;Kalchbrenner, Grefenstette, & Blunsom, 2014;Y. Li, Hao, & Lei, 2016).…”

Section: Methodsmentioning

confidence: 99%

An Analysis of Deep Neural Network for Recommending Developers to Fix Reported Bugs

Abro¹,

Rehman²,

Das³

et al. 2021

European Journal of Science and Technology

View full text Add to dashboard Cite

Occurrence of bugs during the production cycle of software projects is a serious concern of the present time. According to an estimate, a very large number of bugs are recorded while dealing with complex and popular software releases. To locate these bugs and to solve them in efficient manner software industries incorporate the process of bug triage in software testing. Bug triage is intended to recommend the bug reports to an appropriate developer effectively to fix them successfully. However, it becomes labor-intensive and expensive to manually allocate these bug reports to the developer. Deep learning methods have been extensively used and experimented to various domains such as medical diagnosis, earthquake prediction and many more. To handle the above said bugs concerns, many studies have been carried out in order to automate the bug triaging process. Several researchers have directed their efforts by applying deep learning methods in different settings for autonomous recommendation for developers to remove or fix their bugs. In this paper we have proposed a Convolutional Neural Network model for recommending Top 10 developers to fix the reported bugs. For better performance of the model Word2Vec and Glove embeddings are combined with the neural network. The performance of CNN+Word2vec and CNN+Glove models is calculated by averaging the accuracy for 10 developers at five distinct learning rates. The reported results demonstrate that the combination of Convolution with word2vec embedding gives better average accuracy in the testing phase.

show abstract

“…Nearest word in the vocabulary sofa Figure 1: A graphical depiction of the process employed: conversion of the input word into vectors (1), evaluation of the GP tree (2), and conversion of the output vector into a word by finding the most similar word in the vocabulary (3). The fitness can be computed by computing the similarity between the target and output of the GP tree (4) directly.…”

Section: Next Word Prediction With Gpmentioning

confidence: 99%

“…Deep learning (especially recurrent neural models), can capture the sequence information more effectively compared to other existing techniques. The reader is referred to [26] and [2] for a complete review of the existing deep learningbased approaches in the field of NLP.…”

Section: Introductionmentioning

confidence: 99%

Towards an evolutionary-based approach for natural language processing

Manzoni

Jakobović

Mariot

et al. 2020

Proceedings of the 2020 Genetic and Evolutionary Computation Conference

View full text Add to dashboard Cite

Tasks related to Natural Language Processing (NLP) have recently been the focus of a large research endeavor by the machine learning community. The increased interest in this area is mainly due to the success of deep learning methods. Genetic Programming (GP), however, was not under the spotlight with respect to NLP tasks. Here, we propose a first proof-of-concept that combines GP with the well established NLP tool word2vec for the next word prediction task. The main idea is that, once words have been moved into a vector space, traditional GP operators can successfully work on vectors, thus producing meaningful words as the output. To assess the suitability of this approach, we perform an experimental evaluation on a set of existing newspaper headlines. Individuals resulting from this (pre-)training phase can be employed as the initial population in other NLP tasks, like sentence generation, which will be the focus of future investigations, possibly employing adversarial co-evolutionary approaches. CCS CONCEPTS • Computing methodologies → Natural language processing; Genetic programming; Bio-inspired approaches.

show abstract

Improving the Reliability of Deep Neural Networks in NLP: A Review

Cited by 132 publications

References 73 publications

Deep Sentiment Classification and Topic Discovery on Novel Coronavirus or COVID-19 Online Discussions: NLP Using LSTM Recurrent Neural Network Approach

Deep Sentiment Classification and Topic Discovery on Novel Coronavirus or COVID-19 Online Discussions: NLP Using LSTM Recurrent Neural Network Approach

An Analysis of Deep Neural Network for Recommending Developers to Fix Reported Bugs

Towards an evolutionary-based approach for natural language processing

Contact Info

Product

Resources

About