Unsupervised tweets categorization using semantic and statistical features

Article Computer Science, Information Systems

Exploiting dynamic spatio-temporal correlations for citywide traffic flow prediction using attention based neural networks

Ahmad Ali et al.

Summary: DHSTNet, a novel deep spatio-temporal neural network, is proposed for predicting urban traffic crowd flows. The model consists of four components: closeness, period influence, weekly influence, and external branch. Experimental results on real-world datasets demonstrate the superiority of DHSTNet over baseline methods.

INFORMATION SCIENCES (2021)

添加到收藏夹

Article Computer Science, Information Systems

Optimizing Semantic Deep Forest for tweet topic classification

Kheir Eddine Daouadi et al.

Summary: The paper presents a Semantic Deep Forest (SDF) approach that combines Word2vec, WordNet, and Deep Forest for accurate topic detection from Twitter. Experimental results demonstrate that SDF outperforms existing state-of-the-art classification methods in tweet topic classification, without requiring a large amount of labeled data.

INFORMATION SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Interdisciplinary Applications

Low resource language specific pre-processing and features for sentiment analysis task

Loitongbam Sanayai Meetei et al.

Summary: This study conducted sentiment analysis on Manipuri using various machine learning approaches, improving classification results by performing language-specific preprocessing tasks and incorporating additional linguistic features.

LANGUAGE RESOURCES AND EVALUATION (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence