Sentiment Analysis using Multi Head Self-Attention Mechanism Based Bidirectional Gated Recurrent Unit

Published: Aug. 23, 2024

Language: Английский

---

Adaptive Domain-Specific Document-Level Sentiment Analysis with Meta-Learning and Hybrid Language Models

Published: Jan. 1, 2024

Language: Английский

---

Deep Representation Learning of User Preferences for Opinionclassification

Published: Jan. 1, 2023

Text classification assigns predefined categories to text using machine learning models based onlearned patterns on extracted features. Feature-level sparsity occurs as not all features are presentin every sample. This causes misclassifications due improper identification of patterns.Solutions involve feature space reduction and filling. These approaches suffer from informationloss rely domain neighbor-based data combat sparsity. None themhave focused extracting additional contextual information the same textual content toenhance space. We propose a novel deep representation model user preferences tomitigate feature-level by incorporating data. apply this approach in opinionmining product reviews. Further, we employ semantic analysis for interpretationand dimensionality reduction. The proposed method excels over state-of-the-art onseven datasets, surpassing six baseline across diverse metrics. It achieves superior performancethrough combinations methods, with an average reductionexceeding 31% datasets. Compared models, demonstratesimpressive performance enhances multi class accuracy 12%.This has significant potential analysis, achieving high performanceby informative through

Language: Английский

---

Enhanced Word Embedding with CNN using Word Order Information For Sentiment Analysis

2021 18th International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP), Journal Year: 2023, Volume and Issue: unknown, P. 1 - 5

Published: Dec. 15, 2023

On several e-commerce platforms, internet users share opinions. Understanding people's sentiments and opinions is necessary. GloVe, which uses word contexts matrix vectorization, an effective vector learning method. Vector-learning systems have improved with this However, the GloVe model ignores order in context. This paper presents Positional Embedding, integrates into embeddings. We found that our Word Order Vector (PEWOVe) embeddings method outperforms sentiment classification. Amazon data was used to test proposed technique. The 2.7 ^% , 2.5%, 2.1 % more accurate than baseline models on GE, GEC, GE+PE.

Language: Английский

---

Analysis Sentiment towards Delivery Service: Case Study of Paxel

Published: Nov. 7, 2023

Paxel is an application-based delivery service. On Google Play, the application has been downloaded by more than 1 million users and 10 thousand reviews. From existing reviews, it will be useful for company if can processed properly. For this reason, a sentiment analysis was carried out to find sentiments of users, result used as reference improving services or products. The method in research Random Forest, Naive Bayes, Support Vector Machine (SVM). TF-IDF SMOTE data weighting unbalance. As well applying K-fold Cross Validation evaluate used. result, Forest higher accuracy value 91%, Bayes 83%, 87%. Where F1-Score on Positive Class 89%, Neutral 93% Negative 92%.

Language: Английский

---