Data Mining
Fatemeh Akbari; Eynollah Khanjari
Abstract
Background and Objectives: So far, several methods have been proposed to detect communities, which indicate the high importance of discovering communities for understanding social networks and detecting useful and hidden patterns in the network. The goal of such analyses is to find a group of users with ...
Read More
Background and Objectives: So far, several methods have been proposed to detect communities, which indicate the high importance of discovering communities for understanding social networks and detecting useful and hidden patterns in the network. The goal of such analyses is to find a group of users with common characteristics. Basically, social networks are considered as graphs, so the analysis is also done using graph methods, in which nodes represent individuals and edges represent relationships between them. Since community detection is an NP-complete problem, several meta-heuristic approaches have been used to tackle this problem, mainly considering "modularity" as the objective function. In most approaches, modularity has been used, which suffers from the limitation of resolution and cannot detect communities that are small in size and consider it in combination with large communities.Methods: In this paper, a new hybrid algorithm of bee colony and genetics is proposed for community detection which performs optimization using the "balanced modularity" fitness function. In this algorithm, parallel processing is used to speed up optimization, genetic algorithm is used to create the initial population, and genetic operators are used in the search by bees.Results: Experiments on well-known real-world networks, including karate, American football, dolphins, and political books, have shown that our method provides more accurate results than the state-of-the-art community detection methods.Conclusion: The combined optimization of bee colony and genetics not only provides globally optimal solution but also it does not need prior information about the number as well as the structure of communities.
Data Mining
A. Beiranvand; M. Sarhadi; J. Salimi Sartakhti
Abstract
Background and Objectives: Large Language Models have demonstrated exceptional performance across various NLP tasks, especially when fine-tuned for specific applications. Full fine-tuning of large language models requires extensive computational resources, which are often ...
Read More
Background and Objectives: Large Language Models have demonstrated exceptional performance across various NLP tasks, especially when fine-tuned for specific applications. Full fine-tuning of large language models requires extensive computational resources, which are often unavailable in real-world settings. While Low-Rank Adaptation (LoRA) has emerged as a promising solution to mitigate these challenges, its potential remains largely untapped in multi-task scenarios. This study addresses this gap by introducing a novel hybrid approach that combines LoRA with an attention-based mechanism, enabling fine-tuning across tasks while facilitating knowledge sharing to improve generalization and efficiency. This study aims to address this gap by introducing a novel hybrid fine-tuning approach using LoRA for multi-task text classification, with a focus on inter-task knowledge sharing to enhance overall model performance.Methods: We proposed a hybrid fine-tuning method that utilizes LoRA to fine-tune LLMs across multiple tasks simultaneously. By employing an attention mechanism, this approach integrates outputs from various task-specific models, facilitating cross-task knowledge sharing. The attention layer dynamically prioritizes relevant information from different tasks, enabling the model to benefit from complementary insights. Results: The hybrid fine-tuning approach demonstrated significant improvements in accuracy across multiple text classification tasks. On different NLP tasks, the model showed superior generalization and precision compared to conventional single-task LoRA fine-tuning. Additionally, the model exhibited better scalability and computational efficiency, as it required fewer resources to achieve comparable or better performance. Cross-task knowledge sharing through the attention mechanism was found to be a critical factor in achieving these performance gains.Conclusion: The proposed hybrid fine-tuning method enhances the accuracy and efficiency of LLMs in multi-task settings by enabling effective knowledge sharing between tasks. This approach offers a scalable and resource-efficient solution for real-world applications requiring multi-task learning, paving the way for more robust and generalized NLP models.
Data Mining
R. Asgarnezhad; A. Monadjemi; M. SoltanAghaei
Abstract
Background and Objectives: With the extensive web applications, review sentiment classification has attracted increasing interest among text mining works. Traditional approaches did not indicate multiple relationships connecting words while emphasizing the preprocessing phase and data reduction techniques, ...
Read More
Background and Objectives: With the extensive web applications, review sentiment classification has attracted increasing interest among text mining works. Traditional approaches did not indicate multiple relationships connecting words while emphasizing the preprocessing phase and data reduction techniques, making a huge performance difference in classification. Methods: This study suggests a model as an efficient model for sentiment classification combining preprocessing techniques, sampling methods, feature selection methods, and ensemble supervised classification to increase the classification performance. In the feature selection phase of the proposed model, we applied n-grams, which is a computational method, to optimize the feature selection procedure by extracting features based on the relationships of the words. Then, the best-selected feature through the particle swarm optimization algorithm to optimize the feature selection procedure by iteratively trying to improve feature selection. Results: In the experimental study, a comprehensive range of comparative experiments conducted to assess the effectiveness of the proposed model using the best in the literature on Twitter datasets. The highest performance of the proposed model obtains 97.33, 92.61, 97.16, and 96.23% in terms of precision, accuracy, recall, and f-measure, respectively.Conclusion: The proposed model classifies the sentiment of tweets and online reviews through ensemble methods. Besides, two sampling techniques had applied in the preprocessing phase. The results confirmed the superiority of the proposed model over state-of-the-art systems.
Data Mining
Y. Rohani; Z. Torabi; S. Kianian
Abstract
Background: Prediction of students' academic performance is essential for systems emphasizing students' greater success. The results can largely lead to increase in the quality of the educating and learning. Through the application of data mining, useful and innovative patterns can be extracted from ...
Read More
Background: Prediction of students' academic performance is essential for systems emphasizing students' greater success. The results can largely lead to increase in the quality of the educating and learning. Through the application of data mining, useful and innovative patterns can be extracted from the educational data.Methods: In this paper, a new metaheuristic algorithm, combination of simulated annealing and genetic algorithms, is proposed for predicting students’ academic performance in educational data mining. Although metaheuristic algorithms are one of the best options for discovering the hidden relationships between data in data science, they do not separately perform well in accurate prediction of students’ academic performance. Therefore, the proposed method integrates the advantages of both genetic and simulated annealing algorithms. The genetic algorithm is applied to explore new solutions, while simulated annealing is used to increase the exploitation power. By using this combination, the proposed algorithm has been able to predict the students’ academic performance with high accuracy.Results: The efficiency of the proposed algorithm is evaluated on five different educational data sets, including two data sets of students of Shahid Rajaee University of Tehran and three online educational data sets. Our experimental results show and accuracy improvement of the proposed algorithm in comparison to the four similar metaheuristic and five popular classification methods respectively.
Data Mining
R. Asgarnezhad; A. Monadjemi; M. SoltanAghaei
Abstract
Background and Objectives: Twitter Sentiment Classification is one of the most popular fields in information retrieval and text mining. Millions of people of the world intensity use social networks like Twitter. It supports users to publish tweets to tell what they are thinking about topics. There are ...
Read More
Background and Objectives: Twitter Sentiment Classification is one of the most popular fields in information retrieval and text mining. Millions of people of the world intensity use social networks like Twitter. It supports users to publish tweets to tell what they are thinking about topics. There are numerous web sites built on the Internet presenting Twitter. The user can enter a sentiment target and seek for tweets containing positive, negative, or neutral opinions. This is remarkable for consumers to investigate the products before purchase automatically.Methods: This paper suggests a model for sentiment classification. The goal of this model is to investigate what is the role of n-grams and sampling techniques in Sentiment Classification application using an ensemble method on Twitter datasets. Also, it examines both binary and multiple classifications, which are classified datasets into positive, negative, or neutral classes.Results: Twitter Classification is an outstanding problem, which has very few free resources and not available due to modified authorization status. However, all Twitter datasets are not labeled and free, except for our applied dataset. We reveal that the combination of ensemble methods, sampling techniques, and n-grams can improve the accuracy of Twitter Sentiment Classification.Conclusion: The results confirmed the superiority of the proposed model over state-of-the-art systems. The highest results obtained in terms of accuracy, precision, recall, and f-measure.
Data Mining
I. Behravan; S.H. Zahiri; S.M. Razavi; R. Trasarti
Abstract
Background and Objectives: Big data referred to huge datasets with high number of objects and high number of dimensions. Mining and extracting big datasets is beyond the capability of conventional data mining algorithms including clustering algorithms, classification algorithms, feature selection methods ...
Read More
Background and Objectives: Big data referred to huge datasets with high number of objects and high number of dimensions. Mining and extracting big datasets is beyond the capability of conventional data mining algorithms including clustering algorithms, classification algorithms, feature selection methods and etc. Methods: Clustering, which is the process of dividing the data points of a dataset into different groups (clusters) based on their similarities and dissimilarities, is an unsupervised learning method which discovers useful information and hidden patterns from raw data. In this research a new clustering method for big datasets is introduced based on Particle Swarm Optimization (PSO) algorithm. The proposed method is a two-stage algorithm which first searches the solution space for proper number of clusters and then searches to find the position of the centroids. Results: the performance of the proposed method is evaluated on 13 synthetic datasets. Also its performance is compared to X-means through calculating two evaluation metrics: Rand index and NMI index. The results demonstrate the superiority of the proposed method over X-means for all of the synthetic datasets. Furthermore, a biological microarray dataset is used to evaluate the proposed method deeper. Finally, 2 real big mobility datasets, including the trajectories traveled by several cars in the city of Pisa, are analyzed using the proposed clustering method. The first dataset includes the trajectories recorded in Sunday and the second one contains the trajectories recorded in Monday during 5 weeks. The achieved results showed that people choose more diverse destinations in Sunday although it has fewer trajectories. Conclusion: Finding the number of clusters is a big challenge especially fir big datasets. The results achieved for the proposed method showed its fabulous performance in detecting the number of clusters for high dimensional and massive datasets. Also, the results demonstrate the power and effectiveness of the swarm intelligence methods in solving hard and complex optimization problems.======================================================================================================Copyrights©2018 The author(s). This is an open access article distributed under the terms of the Creative Commons Attribution (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, as long as the original authors and source are cited. No permission is required from the authors or the publishers.======================================================================================================
