Urdu Sentiment Analysis via Multimodal Data Mining Based on Deep Learning Algorithms

Sehar, Urooba; Kanwal, Summrina; Dashtipur, Kia; Mir, Usama; Abbasi, Ubaid; Khan, Farzeen

doi:10.1109/access.2021.3122025

Cited by 36 publications

(14 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When learning new tasks, this is mostly due to a lack of effective past knowledge. Literature [7] established the notion of metalearning, which overcomes the aforementioned concerns by accumulating metaknowledge from the learning experience of multiple tasks [8][9][10]. e inductive bias provided by metalearning allows the model to generalize across various tasks, resulting in rapid learning algorithms that are efficient for sampling [10].…”

Section: Metalearningmentioning

confidence: 99%

Integrating Multiclass Light Weighted BiLSTM Model for Classifying Negative Emotions

Bhende¹,

Thakare²,

Pant

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

With the continuous development of social networks, Weibo has become an essential platform for people to share their opinions and feelings in daily life. Analysis of users’ emotional tendencies can be effectively applied to public opinion control, public opinion surveys, and product recommendations. However, the traditional deep learning algorithm often needs a large amount of data to be retained to obtain a better accuracy when faced with new work tasks. Given this situation, a multiclassification method of microblog negative sentiment based on MAML (model-agnostic metalearning) and BiLSTM (bidirectional extended short-term memory network) is proposed to represent the microblog text word vectorization and the combination of MAML and BiLSTM is constructed. The model of BiLSTM realizes the classification of negative emotions on Weibo and updates the parameters through machine gradient descent; the metalearner in MAML calculates the sum of the losses of multiple pieces of training, performs a second gradient descent, and updates the metalearner parameters. The updated metalearner can quickly iterate when faced with a new Weibo negative sentiment classification task. The experimental results show that compared with the prepopular model, on the Weibo negative sentiment dataset, the precision rate, recall rate, and F1 value are increased by 1.68 percentage points, 2.86 percentage points, and 2.27 percentage points, respectively.

show abstract

Section: Metalearningmentioning

confidence: 99%

Integrating Multiclass Light Weighted BiLSTM Model for Classifying Negative Emotions

Bhende¹,

Thakare²,

Pant

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…Tensor Fusion Network (TFN) achieved an accuracy of approximately 74.8 % for textual SA, 69. of Urdu. Built using Bi-LSTM and 3D-CNN approaches, our proposed approach achieved an accuracy of 84 % for unimodal features and 95 % for multimodal features for polarity prediction (Sehar et al, 2021). Despite all this advancement in the development of SA, there is a signi cant research gap in the area of SA for resource-poor languages like Urdu, and the majority of studies also lack concept-level SA.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Chen et al (Chen, Becken, and Stantic, 2019) proposed a lexicon-based approach for SA of Chinese social media posts. They developed a comprehensive process for capturing web posts and a lexicon algorithm for SA in their study (Sehar et al, 2021).…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

A Hybrid Dependency-Based Approach for Urdu Sentiment Analysis

Sehar

Kanwal

Dashtipur

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

In this era of internet, it is evident that social media is one of the biggest platforms acting as a source of producing a huge amount of raw data on a daily basis that contains the opinion of people from different races, cultures, and age groups on a wide range of topics. Data that is produced on the social media platforms could be utilised by businesses to extract information for improving their services and to reach a wider set of audiences based on users' opinions being shared on these social media sites. In order to utilise this huge amount of unstructured data generated continuously on social media, a deep understanding of Natural Language Processing (NLP) is vital. Most of the approaches proposed so far for sentiment analysis (SA) are based on word co-occurrence frequencies that are inefficient in practical cases. Considering this research gap, this paper proposes a framework for concept-level SA for better SA. We developed our Urdu language dataset by collecting from YouTube, which comprises talks and reviews on different topics like movies, political, and commercial products. Further dataset was integrated with language rules and Deep Neural Networks (DNN) to optimise the polarity detection. For SA, when a predefined rule is triggered, the framework enables the sentiments to flow from words to concepts based on dependency relations among different words of a sentence on the basis of Urdu language grammatical rules. When any of the predefined patterns is not triggered, the framework switches to its sub-symbolic counterpart and passes data to DNN for performing the classification of input sentence. The proposed framework outperforms state-of-the-art approaches, including Long short-term memory (LSTM), Convolutional Neural Network (CNN), State Vector Machine (SVM), Linear Regression (LR), and Multilayer Perceptrons (MLP), with a margin of 6–7%, respectively, when employed on our Urdu dataset.

show abstract

“…These varied polarities necessitate that the raw data availabilities be utilized for mining opinions while also identifying their sentiments, as prior literature focuses on textual data, which may fail to generate accurate results [2]. An ideal source of such multimodal information is a video, which provides visual frames and information on spoken language's acoustic and textual representation [2,4]. The integration of these varied data is referred to as Multimodal Sentiment Analysis (MSA).…”

mentioning

confidence: 99%

Fuzzy Layered Convolution Neutral Network for Feature Level Fusion Based On Multimodal Sentiment Classification

Ayodele¹,

Harun²,

Yusoff³

2023

EmAIT

View full text Add to dashboard Cite

Multimodal sentiment analysis (MSA) is one of the core research topics of natural language processing (NLP). MSA has become a challenge for scholars and is equally complicated for an appliance to comprehend. One study that supports MSdifficulties is the MSA, which is learning opinions, emotions, and attitudes in an audio-visual format. In order words, using such diverse modalities to obtain opinions and identify emotions is necessary. Such utilization can be achieved via modality datafusion;such as feature fusion. In handling the data fusion of such diverse modalities while obtaining high performance, a typical machine learning algorithm is Deep Learning (DL), particularly the Convolutional Neutral Network (CNN), which has the capacity to handle tasks of great intricacy and difficulty. In this paper, we present a CNN architecture with an integrated layer via fuzzy methodologies for MSA, a task yet to be explored in improving the accuracy performance of CNN for diverse inputs. Experiments conducted on a benchmark multimodal dataset, MOSI, obtaining 37.5% and 81% on seven (7) class and binary classification respectively, reveals an improved accuracy performance compared with the typical CNN, which acquired 28.9% and 78%, respectively.

show abstract

Urdu Sentiment Analysis via Multimodal Data Mining Based on Deep Learning Algorithms

Cited by 36 publications

References 35 publications

Integrating Multiclass Light Weighted BiLSTM Model for Classifying Negative Emotions

Integrating Multiclass Light Weighted BiLSTM Model for Classifying Negative Emotions

A Hybrid Dependency-Based Approach for Urdu Sentiment Analysis

Fuzzy Layered Convolution Neutral Network for Feature Level Fusion Based On Multimodal Sentiment Classification

Contact Info

Product

Resources

About