LDA and Deep Learning: A Combined Approach for Feature Extraction and Sentiment Analysis

Syamala, Maganti; Nalini, N

doi:10.1109/icccnt45670.2019.8944823

Cited by 5 publications

(4 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ma Haidar et al [27] used LDA model to form document based distribution on the topic of each word, and applied it to the recurrent neural network language model. M. Syamala et al [28] introduced a method to extract the most important information from the views expressed in the input text by combining various machine learning and deep learning techniques with LDA. On the basis of LDA topic model, Niu et al [29] proposed a text classification algorithm by using neural network fitting word topic probability distribution.…”

Section: B Topic Modelmentioning

confidence: 99%

Bi-Level Attention Model With Topic Information for Classification

Liu

Qian

2021

IEEE Access

View full text Add to dashboard Cite

This paper proposes a bi-level attention neural network model (TBAM) that incorporates topic information. This model is suitable for a variety of classification situations and has proven effectiveness in sentiment classification and text classification. Compared with other advanced neural networks that are also used in classification, it has achieved better classification results. The TBAM model has three characteristics: (1) The model reflects the hierarchical structure of the document; (2) The model uses the attention mechanism at the vocabulary and sentence levels respectively, so that it can pay attention to the more important and the less important when constructing the document representation. content. (3) In the vocabulary level, by introducing potential topic information into the semantic representation of the vocabulary level to improve the effect of vocabulary representation, the topic information is regarded as domain information, which makes up for the ignorance of the domain information by the classification model, and increases the model domain adaptability. Through testing on the public data sets CCF-BDCI and THUCNews, the model shows good results on the three indicators of Precision, Recall, and F1.

show abstract

Section: B Topic Modelmentioning

confidence: 99%

Bi-Level Attention Model With Topic Information for Classification

Liu

Qian

2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…And finally used an expectation-maximization algorithm for calculating weightage of each word with respect to its aspect and assigned sentiment. M. Syamala et al; to overcome the limitation of manual topic label assignment to the topics extracted from LDA proposed a deep fusion mechanism [19]. The extracted topics from LDA are converted into word embeddings and trained over a one-layer neural network to determine topic label for each set of extracted topics.…”

Section: A Linguistic Features: Aspect-based Sentiment Analysis Onmentioning

confidence: 99%

An Efficient Aspect based Sentiment Analysis Model by the Hybrid Fusion of Speech and Text Aspects

Syamala¹,

Nalini²

2021

IJACSA

Self Cite

View full text Add to dashboard Cite

Aspect-based Sentiment Analysis (ABSA) is treated to be a challenging task in the domain of speech, as it needs the fusion of acoustic features and Linguistic features for information retrieval and decision making. The existing studies in speech are limited to speech and emotion recognition. The main objective of this work is to combine acoustic features in speech with linguistic features in text for ABSA. A deep learning and language model is implemented for acoustic feature extraction in speech. Different variants of text feature extraction techniques are used for aspect extraction in text. Trained Lexicons, Latent Dirichlet Allocation (LDA) model, Rule based approach and Efficient Named Entity Recognition (E-NER) guided dependency parsing approach has been used for aspect extraction. Sentiment with respect to the extracted aspect is analyzed using Natural Language Processing (NLP) techniques. The experimental results of the proposed model proved the effectiveness of hybrid level fusion by yielding improved results of 5.7% WER and 3% CER when compared with the traditional baseline individual linguistic and acoustic feature models.

show abstract

“…The ML approaches such as Support Vectors Machines (SVM) combined with Maximum Entropy, Naive Bayes (NB) [5][6][7], and Latent Dirichlet Allocation (LDA) were proposed in recent years and trained based on sentiment analysis [8][9]. The ML approaches are advanced into deep learning models, Various Deep learning models are Convolutions Neural Networks (CNN), Deep Neural Network (DNN), Deep Restricted Boltzmann's Machine (RBM), Deep CNN, etc.…”

Section: Introductionmentioning

confidence: 99%

A Novel Deep RCNN-DCLSTM Classifier for Finding the Quality of Web-Page Based On Review Dataset

Srivast,

Pandey,

Agarwal

2023

Preprint

View full text Add to dashboard Cite

Nowadays, internet has numerous of web contents but it is difficult to find the web page quality. For predicting the quality of web page, a technique is necessary. Therefore, for determining the quality of the web page, a novel Residual Convolutional Neural network and Drop Connect Long Short Term Memory (RCNN-DCLSTM) technique is proposed. It consists of two stages: pre-processing and classification. In the preprocessing stage, tokenization, identification of slang, stop word removal, and lemmatization processes improve the level of accuracy during classification. In the classification stage, the proposed deep learning classifier based on RCNN-DCLSTM is used to classify the quality of web page as very high quality, high quality, moderate quality, low quality, and very low quality based on reviews. Here, the Drop connect regulation system on hidden-to-hidden weight metrics with LSTM is used to avoid the fitting problem. The proposed RCNN-DCLSTM accuracy is tested on four data sets and compared with previous methods. Based on the estimation result, the proposed RCNN-DCLSTM gives the accuracy of 0.91, precision of 0.909, recall of 0.908, and F-1 measure of 0.91. Hence, it is proved that the proposed RCNN-DCLSTM technique accurately finds the quality of web page.

show abstract

LDA and Deep Learning: A Combined Approach for Feature Extraction and Sentiment Analysis

Cited by 5 publications

References 15 publications

Bi-Level Attention Model With Topic Information for Classification

Bi-Level Attention Model With Topic Information for Classification

An Efficient Aspect based Sentiment Analysis Model by the Hybrid Fusion of Speech and Text Aspects

A Novel Deep RCNN-DCLSTM Classifier for Finding the Quality of Web-Page Based On Review Dataset

Contact Info

Product

Resources

About