Skip to main content

IJICTR Journal

ICT Research Institute, ICT Security Faculty, Faculty Member

ICT Research Institute, Ict, Faculty Member

Followers

220

Following

283

Co-authors

51

Public Views

International Journal of Information and Communication Technology Research (IJICTR) publishes high quality scientific papers covering all aspects of information and communication technologies. IJICTR is an open access journal which publishes research across all academic disciplines related to ICT. It is published by ICT Research Institute.
Phone: +982188630077
Address: International Journal of Information and Communication Technology Research, ICT Research Institute, Tehran, Iran.

less

Interests

Uploads

Papers by IJICTR Journal

Investigating Sentiment Analysis of News in Stock Market Prediction Utilizing Machine Learning Technique

International Journal of Information and Communication Technology Research, 2024

In the stock market, which is a dynamic, complex, nonlinear and non-parametric environment, accur... more In the stock market, which is a dynamic, complex, nonlinear and non-parametric environment, accurate prediction is crucial for trading strategy. It is assumed that news articles affect the stock market. We investigated the relationship[1] between headline’s sentiment of news and their impact on stock prices changes. To show this relationship, we applied the sentiment data and the price difference between the day before the news was published and the day of the news, to machine learning regression and classification models. Regression is used to predict changes and classification is used to decide whether to buy or sell stocks. We used three stock datasets named Apple, Amazon and AXP and the results are shown in the mentioned dataset that using news with negative sentiments can make predictions just as correctly as using news with both positive and negative sentiments. In regression and classification models, Random Forest outperformed other machine learning algorithms in predicting stock price changes using news sentiment analysis. Additionally, we depicted that the results of computer and human tagging were almost similar, showing that using computer tools for text tagging will allow to tag text much more quickly and easily.

A Deep Learning Approach for Sarcasm Detection on Twitter

International Journal of Information and Communication Technology Research, 2024

Sarcasm is a form of speech in which a person expresses his opinion implicitly. We may encounter ... more Sarcasm is a form of speech in which a person expresses his opinion implicitly. We may encounter a seemingly positive sentence in sarcasm, but the speaker has a contrary opinion. Sarcasm can be recognized in spoken language based on body language and the tone of voice. However, the lack of these features makes it difficult to recognize sarcasm in text. In recent years, Twitter has attracted much attention and has become a popular platform for sharing opinions and viewpoints. It is also common for people to use sarcasm on Twitter as an indirect means of expressing their opinions. The presence of sarcasm in the text makes it difficult to recognize the sentiment. Thus, it is necessary and inevitable to have solutions that can detect sarcasm. This study aims to provide a solution for detecting sarcasm on Twitter using deep learning approaches. This study used two Twitter datasets containing balance and imbalance data for modeling. The main idea of this research is to use additional features such as sentimental features, subjectivity, number of hashtags, and punctuation along with features that deep learning algorithms automatically extract. The impact of each feature is reported in the paper. In this research, GRU-Capsule based neural network has been used. According to the results, the proposed model has improved accuracy by 5% for balanced data and by 2% for imbalanced data

EmoRecBiGRU: Emotion Recognition in Persian Tweets with a Transformer-based Model, Enhanced by Bidirectional GRU

International Journal of Information and Communication Technology Research, 2024

Emotion recognition in text is a fundamental aspect of natural language understanding, with signi... more Emotion recognition in text is a fundamental aspect of natural language understanding, with significant applications in various domains such as mental health monitoring, customer feedback analysis, content recommendation systems, and chatbots. In this paper, we present a hybrid model for predicting the presence of six emotions: anger, disgust, fear, sadness, happiness, and surprise in Persian text. We also predict the primary emotion in the given text, including these six emotions and the “other” category. Our approach involves the utilization of XLM-RoBERTa, a pre-trained transformer-based language model, and fine-tuning it on two diverse datasets: EmoPars and ArmanEmo. Central to our approach is incorporating a single Bidirectional Gated Recurrent Unit (BiGRU), placed before the final fully connected layer. This strategic integration empowers our model to capture contextual dependencies more effectively, resulting in an improved F-score after adding this BiGRU layer. This enhanced model achieved a 2% improvement in the F-score metric on the ArmanEmo test set and a 7% improvement in the F-score metric for predicting the presence of six emotions on the final test set of the ParsiAzma Emotion Recognition competition.

Claim Detection in Persian Twitter Posts

International Journal of Information and Communication Technology Research, 2024

The proliferation of false information on social media has profound negative impacts across vario... more The proliferation of false information on social media has profound negative impacts across various aspects of people's lives. To mitigate these effects, numerous studies have focused on developing automated factchecking systems aimed at enhancing the accuracy and reliability of news and information. Claim detection, recognized as the initial stage in constructing such systems, has been explored in several languages. In our paper, we introduce a corpus of Persian tweets annotated with 11 labels derived from linguistic analysis, representing different types of claims. Additionally, we establish a baseline claim detection model to assess the dataset. This study frames claim detection as a classification task and employs a transformer-based approach to train a multi-label classifier capable of identifying various types of claims in Persian texts.

An Improvement in Transformer-Based Sentiment Analysis in Persian Twitter

International Journal of Information and Communication Technology Research, 2024

In the era of deep learning, transformer-based models have revolutionized natural language proces... more In the era of deep learning, transformer-based models have revolutionized natural language processing tasks, offering unparalleled performance in capturing contextual relationships. This paper delves into the realm of sentiment analysis in Persian Twitter, employing state-of-the-art transformer architectures. Through rigorous experimentation on a dedicated Persian sentiment dataset, we explore the capabilities of transformers in deciphering nuanced emotions expressed in tweets. The results demonstrate the potency of these models, highlighting their effectiveness in understanding the intricacies of sentiment within the Persian language. This study not only contributes insights into sentiment analysis but also underscores the transformative impact of transformer architectures in unlocking the expressive dynamics of Persian social media discourse. We trained multiple deep learning architectures based on transformers for sentiment analysis on Persian Twitter data, and in the test section, we achieved a 60.37% F-score.

A Comparative Study of BERT-X for Sentiment Analysis and Stance Detection in Persian Social Media

by IJICTR Journal, Mohamad Sobhi, and Alireza Mazochi

International Journal of Information and Communication Technology Research, 2024

BERT-based models have gained popularity for addressing various NLP tasks, yet the optimal utiliz... more BERT-based models have gained popularity for addressing various NLP tasks, yet the optimal utilization of knowledge embedded in distinct layers of BERT remains an open question. In this paper, we introduce and compare diverse architectures that integrate the hidden layers of BERT for text classification tasks, with a specific focus on Persian social media. We conduct sentiment analysis and stance detection on Persian tweet datasets. This work represents the first investigation into the impact of various neural network architectures on combinations of BERT hidden layers for Persian text classification. The experimental results demonstrate that our proposed approaches can outperform the vanilla BERT that utilizes an MLP classifier on top of the corresponding output of the CLS token in terms of performance and generalization.

ParsiAzma Challenges on Persian Text Analysis in Social Media

by IJICTR Journal and Maryam Mahmoudi

International Journal of Information and Communication Technology Research, 2024

The ParsiAzma[1] challenges in 2023 focused on Improving Persian text analysis in social media. W... more The ParsiAzma[1] challenges in 2023 focused on Improving Persian text analysis in social media. We designed four shared tasks: stance detection, sentiment analysis, emotion detection, and claim detection in social media posts. The goal of these challenges was to bring together various teams to develop the best models for these challenges and to establish a standard test platform for future Persian language research. A total of 28 teams participated, competing to solve the specified tasks. The most effective models in all shared tasks utilized the BERT model. Text embedding was first obtained using a BERT[2]-based model, followed by final predictions with either an MLP[3] or CNN[4]. Additionally, several meta-classifiers were developed as fusion models to leverage the strengths of individual models. The best results based on accuracy criteria for the four challenges—stance detection, sentiment analysis, emotion recognition, and claim detection—were 0.67, 0.67, 0.45, and 0.56, respectively. These results indicate that emotion detection has lower accuracy than the other three tasks, highlighting its complexity.

A Hybrid Model for Evaluation of E-Learning User Satisfaction with TAM and ELQ Approach

International Journal of Information and Communication Technology Research, 2024

With the start of the Coronavirus epidemic, remote education systems were widely accepted in all ... more With the start of the Coronavirus epidemic, remote education systems were widely accepted in all universities around the world. The University of Tehran, the top university in Iran, had an electronic education system since 2010, which provided services in this field to the academic community. With the spread of the pandemic, these services became widespread in all departments of the University of Tehran. Compared to traditional learning, E-Learning brings challenges for Users, the main challenge of which is Users’ Satisfaction. In this paper, by examining the developed model based on the TAM model and the factor analysis of this model, the level of satisfaction of student users with the elearning management system of the University of Tehran during the Corona era is carefully evaluated. The exploitation of this model with the analytical method of partial least squares in the construction of structural equations is based on the information collected from the questionnaire that has been provided to the users of the e-learning system of the University of Tehran. The results obtained from this study show that the Technology Acceptance Model (TAM) describes well the factors affecting the level of user satisfaction with the E-learning management system. The positive effect of the visual beauty and the quality of the information in this system has improved the technical quality of the service; as a result, the level of user satisfaction in using this system has increased.

Image Retrieval with Missing Regions by Reconstructing and Concatenating Content and Semantic Features

International Journal of Information and Communication Technology Research, 2024

In recent years, the performance of deep neural networks in improving the image retrieval process... more In recent years, the performance of deep neural networks in improving the image retrieval process has been remarkable. Utilizing deep neural networks; however, leads to poor results in retrieving images with missing regions. The operators’ dysfunctions, who consider the relationship between the image pixels, statistically extract incomplete information from an image, which in turn reduces the number of image features and or leads to features' inaccurate identification. An attempt has been made to eliminate the problem of missing image information through image inpainting techniques; therefore, a content-based image retrieval method is proposed for images with missing regions. In this method, through image inpainting the crucial missing information is reconstructed. The image dataset is being queried to find similar samples. For this purpose, a two-stage inpainting framework based on encoder-decoder is used in the image retrieval system. Also, the features of each image are extracted from the integration and concatenating of content and semantic features. Through using handcraft features such as color and texture image content information is extracted from the Resnet-50 deep neural network. Finally, similar images are retrieved based on the minimum Euclidean distance. The performance of the image retrieval model with missing regions is evaluated with the average precision criterion on the Paris 6K datasets. The best retrieval results are 60.11%, 50.14%, and 42.43% for retrieving the top one, five, and ten samples after reconstructing the image with the most missing regions with a destruction frequency of 6 Hz, respectively.

Deep Reinforcement Learning-based Exploration of Web Applications

International Journal of Information and Communication Technology Research, 2024

—Web application (app) exploration is a crucial part of various analysis and testing techniques. ... more —Web application (app) exploration is a crucial part of various analysis and testing techniques. However, the current methods are not able to properly explore the state space of web apps. As a result, techniques must be developed to guide the exploration in order to get acceptable functionality coverage for web apps. Reinforcement Learning (RL) is a machine learning method in which the best way to do a task is learned through trial and error, with the help of positive or negative rewards, instead of direct supervision. Deep RL is a recent expansion of RL that makes use of neural networks’ learning capabilities. This feature makes Deep RL suitable for exploring the complex state space of web apps. However, current methods provide fundamental RL. In this research, we offer DeepEx, a Deep RL-based exploration strategy for systematically exploring web apps. Empirically evaluated on seven open-source web apps, DeepEx demonstrated a 17% improvement in code coverage and a 16% enhancement in navigational diversity over the stateof-the-art RL-based method. Additionally, it showed a 19% increase in structural diversity. These results confirm the superiority of Deep RL over traditional RL methods in web app exploration.

An Overview of Secure Communications for the Internet of Things

International Journal of Information and Communication Technology Research, 2024

As an emerging technology that combines both digital and physical realms, access to information t... more As an emerging technology that combines both digital and physical realms, access to information technology has expanded (IoT) the Internet of Things. The Internet of Things, as it becomes more pervasive, will overshadow human life as much as possible. Some of the major challenges associated with the development of this phenomenon have been the issue of security, which is needed in all its layers and even specifically in individual layers. According to the structure and applications of the Internet of Things, as well as the threats and challenges in cyberspace, we first examine security needs and then, by examining some methods of securing the Internet of Things, we propose a method according to the approaches discussed.

A Routing Method in Smart Cities Based on Destination Type and Network Traffic Flow

by IJICTR Journal and Mohammad Moradi

International Journal of Information and Communication Technology Research, 2024

The optimal routing will only be possible if the link weights are calculated correctly. Also, the... more The optimal routing will only be possible if the link weights are calculated correctly. Also, the weight of the links should be different based on the type of destination nodes and the traffic flow of each path in the network and should be updated in real time. In this study, using multi-criteria decision-making techniques, the criteria related to each type of destination node and the importance of that criterion according to the type of the destination node, are considered. This research is applied based on the purpose. Library studies have been used to extract criteria related to each type of the destination node. Also, the field method has been used to calculate the weight of each criterion. The simulation of the proposed method showed that this method has reduced the average weight of links in routing by about half compared to the method that considered only the "distance" criterion.

Optimal Power Allocation Strategy for a D2D Network based Safety-Critical V2X Communication with Partial CSI

International Journal of Information and Communication Technology Research, 2024

—In intelligent transport system (ITS) networks, it is very useful to reduce the transmission ene... more —In intelligent transport system (ITS) networks, it is very useful to reduce the transmission energy consumption. Device-to-device (D2D) communication is a key technology that can improve the performance of the ITS networks. In this paper, we investigate a D2D-enabled vehicular network and study a power allocation strategy to maximize the spectral efficiency of the network. Also, for safety-critical vehicle-to-vehicle (V2V) and vehicle-toinfrastructure (V2I) applications, it is presented a reliable connection during the V2V and V2I links. The performance analysis of the investigated network shows that the studied method can maximize the total spectral efficiency of the network subject to satisfy some constraints when the channel state information (CSI) is perfectly known or not.

Introduce an Enhanced Hospital Information System Reference Architecture with ATAM Evaluation

International Journal of Information and Communication Technology Research , 2024

—Hospital information system, as an integrated system consists of complex, diverse, and heterogen... more —Hospital information system, as an integrated system consists of complex, diverse, and heterogeneous information systems, which supports all clinical, administrative, and financial activities of patients; therefore, it faces interoperability problems, which is a critical feature of data sharing and integration. In this regard, a reference architecture will be presented for the integration of different systems in the hospital which comprises system, information and software architecture. This architecture emphasizes interoperability with eight layers of user interface and application, services, data collection and storage, integration, external application systems, communication and information infrastructure, management, and security/privacy, based on service orientation. It will be evaluated by ATAM method, based on ten scenarios. It shows the presented architecture will satisfy all functional and non-functional requirements such as interoperability between different information systems, reduction of information redundancy and development cost and time, scalability and accessibility.

An Optimized Deep Model for Bi-lingual Sentiment Analysis of Medical User Reviews

International Journal of Information and Communication Technology Research , 2024

—Sentiment analysis of online doctor reviews helps patients to better evaluate and select the rel... more —Sentiment analysis of online doctor reviews helps patients to better evaluate and select the related doctors based on the previous patients' satisfaction. Although some studies are addressing this problem in the English language, only one preliminary study has been reported for the Persian language. In this study, we propose a new evolutionary deep model for sentiment analysis of Persian online doctor reviews. The proposed method utilizes both Persian reviews and their English translations as inputs of two separate deep models. Then, the outputs of the two models are combined in a single vector which is used for deciding the sentiment polarity of the review in the last layer of the proposed deep model. To improve the performance of the system, we propose an evolutionary approach to optimize the hyperparameters of the proposed deep model. We also compared three evolutionary algorithms, namely, Ant Colony Optimization (ACO), Genetic Algorithm (GA), and Gray Wolf Optimization (GWO) algorithm, for this purpose. We evaluated the proposed model in two phases; In the first phase, we compared four deep models, namely, long shortterm memory (LSTM), convolutional neural network (CNN), a hybrid of LSTM and CNN, and a bidirectional LSTM (BiLSTM) model with four traditional machine learning models including Naïve Bayes (NB), decision tree (DT), support vector machines (SVM), and random forest (RF). The results showed that the BiLSTM and CNN models outperform other methods, significantly. In the second phase, we compared the optimized version of two proposed bi-lingual models in which either two BiLSTM or two CNN models were used in parallel for processing Persian and English reviews. The results indicated that the optimization of the CNN using ACO and the optimization of BiLSTM using a genetic algorithm can achieve the best performance among other combinations of two deep models and three optimization algorithms. In the current study, we proposed two deep models for bi-lingual sentiment analysis of online Persian doctor reviews. Moreover, we optimized the proposed models using ACO, genetic algorithm, and gray wolf optimization methods. The results indicated that the proposed bi-lingual model outperforms a similar model using only Persian or English reviews. Also, optimizing the parameter of proposed deep models using ACO or genetic algorithms improved the performance of the models.

Software Re-Modularization Method Based on Many-Objective Function

International Journal of Information and Communication Technology Research , 2024

Software evolution and continuous changes make maintenance difficult, reducing the quality of sof... more Software evolution and continuous changes make maintenance difficult, reducing the quality of software structure and architecture. To cope with this challenge, re-modularization is used to promote the modular structure of software system by the re-grouping of software elements. In this paper, the proposed method recognizes various dependencies in terms of an objective function unlike what has been stated in some other methods. In this method, a search-based many-objective fitness function is proposed to formulate re-modularization as an optimization problem. The results of the proposed method have been compared to the effects of four other methods based on MQ and NED. The results show the proposed method improved re-modularization remarkably compared to others in terms of both MQ and NED criteria especially for smaller software. Therefore, the proposed method can be effective in redefining real-world applications.

Reducing The Computational Complexity of Fuzzy Identity-Based Encryption from Lattice

International Journal of Information and Communication Technology Research , 2024

—In order to provide access control on encrypted data, Attribute-based encryption (ABE) defines e... more —In order to provide access control on encrypted data, Attribute-based encryption (ABE) defines each user using a set of attributes. Fuzzy identity-based encryption (FIBE) is a variant of ABE that allows for a threshold access structure for users. To address the potential threat posed by future quantum computers, this paper presents a postquantum fuzzy IBE scheme based on lattices. However, current lattice-based ABE schemes face challenges related to computational complexity and the length of ciphertext and keys. This paper aims to improve the performance of an existing fuzzy IBE scheme by reducing key length and computational complexity during the encryption phase. While negative attributes are not utilized in our scheme, we prove its security under the learning with error (LWE) hard problem assumption in the selective security model. These improvements have significant implications for the field of ABE.

Energy Efficient Distributed Anomaly Detection using Semi-Supervised Models in IoT

International Journal of Information and Communication Technology Research , 2024

Utilizing IoT technologies for monitoring large-scale smart facilities such as power, water and g... more Utilizing IoT technologies for monitoring large-scale smart facilities such as power, water and gas distribution networks has been the subject of many studies recently. The aim is to detect anomalous events in the network due to elements’ failure, bad designs, attacks or abuses of the network and alert the network operators in a timely manner. As the centralized cloud-based approaches are impractical in time-critical and real-time anomaly detection applications due to 1) high sensor-to-cloud transmission latency 2) high communication cost and 3) high energy consumption at the sensor nodes, the distributed anomaly detection methods based on Deep Neural Networks (DNN) have been applied in past studies vastly. In these methods, in order to detect anomalies in real-time, copies of the anomaly detection model are placed at the sensor nodes (rather than placing one at the cloud node) reducing the sensor-to-cloud transmissions significantly. Nevertheless, new normal samples collected at the sensor nodes still need to be transmitted to the cloud node at predefined intervals to re-train the distributed anomaly detection DNNs. In order to minimize these sensor-to-cloud transmissions during the retraining process, in this paper, two well-known lossless coding algorithms: Huffman Coding and Arithmetic Coding were studied and it was observed that the Huffman and Arithmetic Coding were able to reduce the transmission traffic up to 50% and 75% respectively using two IoT benchmark datasets of pipeline measurements. Besides, the Huffman Coding shown to be computationally feasible on resource limited sensors and resulted in up to 10% saving in energy consumption on each sensor resulting in longer network longevity. Moreover, the experimental results showed that the auto-encoder DNN could outperform the one-class SVM in the iterative distributed anomaly detection method.

A Dynamic Hybrid Precoding Structure for mmWave Massive MIMO Systems

by IJICTR Journal and Amirreza Moradi

International Journal of Information and Communication Technology Research , 2024

Due to the high-power consumption and complexity of fully digital baseband precoding, its impleme... more Due to the high-power consumption and complexity of fully digital baseband precoding, its implementation in massive millimeter-wave multiple-input multiple-output (MIMO) systems is not cost-efficient and practical; for this reason, hybrid precoding has attracted a lot of attention in recent years. Most hybrid precoding techniques concentrate on the fully-connected structure, although they require lots of phase shifters, which is high energy-consuming. On the contrary, the partially-connected structure has low power consumption, nevertheless, suffers from a severe decrease in spectral efficiency (SE). To enhance SE, this paper proposed a dynamic hybrid precoding structure where a switch network is able to provide dynamic connections from phase shifters to radio frequency (RF) chains. To determine the digital precoder and the states of switch, a novel alternating minimization algorithm is proposed, which leverages closed-form solutions at each iteration to efficiently converge to an optimal solution. Furthermore, the phase shifter matrix is optimized through an iterative solution. The simulation results show that in terms of SE, the proposed algorithm with a dynamic structure achieves higher performance than the partial structure. Also, since the proposed structure reduces the number of phase shifters, it can guarantee better energy efficiency (EE) than the fully connected structure

Resource Allocation optimization in fog Architecture Based Software-Defined Networks

International Journal of Information and Communication Technology Research , 2023

As a growing of IoT devices, new computing paradigms such as fog computing are emerging. Fog comp... more As a growing of IoT devices, new computing paradigms such as fog computing are emerging. Fog computing is more suitable for real-time processing due to the proximity of resources to IoT layer devices. Service providers must dynamically update the hardware and software parameters of the network infrastructure. Software defined network (SDN) proposed as a new network paradigm, whose separate control layer from data layer and provides flexible network management. This paper presents a software-defined fog platform to host real-time applications in IoT. Then, we propose a novel resource allocation method. This method involves scheduling multi-node real-time task graphs over the fog to minimize task execution latency. The proposed method is designed to benefit the centralized structure of SDN. The simulation results show that the proposed method can find near to optimal solutions in a very lower execution time than the brute force method.