Vietnam Journal of Computer Science

Community detection is a highly active research area that aims to identify groups of vertices with similar properties or interests within complex real-world networks. Over the years, a large number of papers have been published, resulting in the development of various community detection algorithms that consider factors such as the type of networks, the nature of communities, and applications. Despite numerous relevant review and survey papers, the literature lacks a comprehensive analysis of the computational complexity of existing community detection algorithms. This review aims to address this gap by providing a more detailed analysis and evaluation of the computational complexity of hierarchical clustering algorithms for community detection, including two main categories: agglomerative and divisive algorithms. We also highlight the main theoretical concepts, emphasizing the benefits and drawbacks of each approach, both in theory and in practical applications. This review helps researchers and practitioners in this field better understand valuable information on the differences and unique features of community detection algorithms with hierarchical community structure.

This paper presents an approach to improve the file fragment classification by proposing new features for classification and evaluating them on a dataset that includes both low- and high-entropy file fragments. High-entropy fragments, belonging to compressed and encrypted files, are particularly challenging to classify because they lack exploitable patterns. To address this challenge, the proposed feature vectors are constructed based on the byte frequency distribution (BFD) of file fragments, along with discrete Fourier transform coefficients and several randomness measures. These feature vectors are tested using three machine learning models: Support vector machines (SVMs), artificial neural networks (ANNs), and random forests (RFs). The proposed approach is evaluated on the govdocs1 dataset, which is freely available and widely used in this field, to enable reproducibility and fair comparison with other published research. The results show that the proposed approach outperforms existing methods and achieves better classification accuracy for both low- and high-entropy file fragments.

In accrediting training quality, assessing the educational quality of a university is an extremely important task to affirm the prestige of the university to the society. Areas of assessment include training program, teaching and learning activities, capacity of graduate student, and the reputation of the university in society. This research studies the characteristics of the survey dataset structures to form data variables and analyzes the relations between data variables to build non-spatial and temporal multidimensional model (nSTM), applies the establishment of visual graphs according to the data variables representing the datasets with vertical bars indicating the values of the data variables being visually represented. This research conducts experiments on the dataset structures surveying the areas of a university. With this scientific approach, it will help university administrators to build policies to improve in all aspects and especially improve the quality of university education.

Sentiment analysis (SA), also known as opinion mining, is a natural language processing (NLP) technique used to determine the sentiment or emotional tone behind a piece of text. It involves analyzing the text to identify whether it expresses a positive, negative, or neutral sentiment. SA can be applied to various types of text data such as social media posts, customer reviews, news articles, and more. This experiment is based on the Internet Movie Database (IMDB) dataset, which comprises movie reviews and the positive or negative labels related to them. Our research experiment’s objective is to identify the model with the best accuracy and the most generality. Text preprocessing is the first and most critical phase in an NLP system since it significantly impacts the overall accuracy of the classification algorithms. The experiment implements unsupervised sentiment classification algorithms including Valence Aware Dictionary and sentiment Reasoner (VADER) and TextBlob. We also examine the supervised sentiment classifications methods such as Naïve Bayes (Bernoulli NB and Multinomial NB). The Term Frequency-Inverse Document Frequency (TFIDF) model is used to feature selection and extractions. The combination of Multinomial NB and TFIDF achieves the highest accuracy, 87.63%, for both classification reports based on our experiment result.

Lungs are crucial parts of the human body and can be captured as Chest x-ray images for disease diagnosis. Unfortunately, in many countries, hospitals and healthcare centers lack qualified doctors for medical images-based diagnosis. Recent numerous advancements in artificial intelligence have deployed with many medical applications to support doctors for disease diagnosis. In our research, we have leveraged YOLOv5s to identify and extract lungs and performed segmentation tasks with Fast R-CNN and YOLOv5 for comparison. The lung region abnormality detection models have pretty good average precision. For example, the YOLOv5 model outperforms both in terms of training time, prediction, and accuracy, with the AP@.5 and AP@.5:.95 metric values, 0.616 and 0.322 on 2,500 images of 5 abnormalities (aortic enlargement, cardiomegaly, lung opacity, pleural effusion, and pulmonary fibrosis).

Item−item collaborative filtering is a sub-type of a recommender system that applies the items’ similarities for recommending a new set of items to the user. Usually, a traditional recommender system utilizes items’ ratings given by the user for deducing their preferences for recommending items. However, for the popularity of social platforms, users are now more familiar to write textual comments known as reviews about items based on their experiences rather than giving a rating, because rating any item limits a user to manifest the degree of satisfaction towards the item. As a result, the items’ reviews become a precious source of information that could enhance the system’s performance. In this paper, a novel recommendation approach has been proposed by applying a recurrent neural network to incorporate items’ reviews with the recommender system. The recurrent neural network is a deep learning-based approach that can distribute the text to the relevant classes. Thus, the proposed approach has applied long short-term memory which is a modern formation of recurrent neural network that is applied to compute items’ rating scores from the items’ reviews. Then, the score is used to define the uniformity of items by using the Jaccard and Pearson correlation coefficient. The proposed approach has been evaluated by two familiar datasets named Yelp & Amazon datasets. Also, it is found that the proposed approach surpasses the traditional techniques and also improved the accuracy of prediction for the Yelp dataset by approximately in respect of 1.37% mean absolute error, 2.17% precision, 2.08% recall, and 2.11% f-measure. Furthermore, the proposed approach increased the recommendation performance for Amazon dataset on average in term of 1.34% mean absolute error, 2.09% precision, 2.53% recall, and 2.32% f-measure, respectively.

In this paper, we consider the issue of digital camera identification based on images. This topic matches the area of digital forensics. The problem is well known in the literature and many algorithms based on camera’s fingerprint have been proposed. In this paper, we discuss the digital camera identification based on convolutional neural networks (CNN). CNNs are state-of-the-art method in computer vision and are widely utilized in many applications. Our goal is to find out whether it is possible to speed up the process of learning the networks by the images. We conduct a set of representative experiments which show that replacing the ReLU with SELU activation function and adjusting the network’s hyperparamethers (e.g. learning rate) may have a significant impact on reduction time of learning. We also consider using the dropout layer. The experiments are held on representative image dataset, consisting of many images coming from modern cameras and show effectiveness of our propositions.