Shahgholi, B.,
Shahgholi ghahfarokhi, B.,
Shahbazi, H.,
Kazemifard M.,
Zamanifar, K. Publication Date: 2006
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 74-80
RoboCupRescue Simulation System is a platform for designing and implementing various artificial intelligent issues. In rescue simulation environments, Firebrigades should select fire points in a collaborative manner such that the total achieved result is optimized. In this work, we are going to propose a new method for fire prediction and selection in Firebrigade agents. This method is based on Evolving Fuzzy Neural Networks to obtain a set of trained fuzzy rules as rule base of Firebrigades Fire Selection System to select targets autonomously.
Mahdavi, M.,
Arasteh, S.,
Mahdavi, M.,
Bideh, P.N.,
Hosseini S.,
Chapnevis A. Publication Date: 2018
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 1499-1504
Watermarking refers to a way of protecting copyright. From one perspective, watermarking algorithms are categorized as key-based and non-key-based. In the key-based watermarking, targeted blocks are selected by using a key to embed the watermark bits; in non-key-based watermarking on the other hand, watermark bits are embedded in a predefined position of the host image. In general, key-based schemes are expected to be more robust against targeted removal attacks. In this paper we propose two attacks against two key-based watermarking algorithm, which are based on the QR decomposition. Although the key based watermarking algorithms are supposed to be more robust, we show that they are vulnerable to our proposed attacks. Our experimental results indicate that the existence of a key is also not a guarantee for the safety of watermarking methods. © 2018 IEEE.
Publication Date: 2017
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 1469-1474
Covert channels have attracted serious attention in network and computer security since their first presentation by Lampson in 1973. Designing such channels can be seen either as a threat or an opportunity, especially in ad-hoc networks. Lots of opportunities were studied to implement effective covert channels in different layers of the network. However, most of the proposed covert channels are designed in MAC and network layers, in ad-hoc networks. In this paper, we want to establish a covert channel on the base of ExOR, a popular opportunistic routing protocol in ad-hoc networks. The proposed channel connects the covert parties in the middle of the existing overt communication. This method transfers covert bits by taking the control of the number of packets in the covert sender's fragment. The performance and detectability of the method is evaluated in this paper and the author's direction of future study is also discussed. © 2017 IEEE.
Publication Date: 2016
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 32-37
The concept of digital watermarking has been developed to solve problems such as illegal duplication and distribution of digital media. In watermarking, the process of removing the watermark from the media is known as an attack. Typically, attacks are carried out using tools such as Stirmark. Some attacks are executed in a targeted manner; in other words, knowing the watermarking algorithm, they directly seek to destroy the watermark in the media. In this type of attack, the damage caused to the media is less extensive than generalized attacks such as Stirmark. Clearly, targeted attacks require prior knowledge about the watermarking algorithm. To the best of our knowledge, algorithm detection in watermarking remains to be investigated. One possible approach is to use staganalysis feature sets; however, we demonstrate that, despite their large number of features, such feature sets do not produce adequate results for watermarking. In this paper, several features are introduced, which can be used in an SVM classifier to allow the detection of the watermarking algorithm. According to implementation results, although the proposed feature set is small, its accuracy is substantially greater than that of the staganalysis feature sets. © 2016 IEEE.
Publication Date: 2016
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 60-65
In these days, location based services are being more popular than ever and can be seen in ways like simple reservation systems to more complex commercial applications with their customized e-commerce logic. All of these applications are based on sending users locations and other confidential data over untrusted channels which can expose such private data to suicide hackers eavesdropping communication channel. One of these applications that this paper tries to cover is location based mobile coupons. In this service, users receive mobile coupons based on their location information from nearby stores. Sending location information to service providers without any consideration can disclose user location to malignant users or even service provides can use these information to track the user. In this paper we use anonymous authentication to preserving location privacy using blind signature. The unforgeability and unlinkability features of proposed method avoid coupon frauds and location tracking together. © 2016 IEEE.
Publication Date: 2016
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 13-19
Semi-fragile watermarking methods are mainly characterized by their robustness against unintentional attacks such as regular image processing and their fragility against intentional attacks aiming to alter signal contents. Thus, these methods need to provide correct robustness and fragility. However, numerous factors impede the achievement of these two objectives. The fact that the watermark is not dependent on the content of the image (the medium considered in this study) and also the semi-fragile method is not block-wise dependent, make the method vulnerable to intentional attacks. Recently, an image authentication method has been proposed, which despite the use of a secret key in generating the watermark, suffers from the preceding vulnerabilities. Thus, in this paper we aim to present a counterfeiting attack and demonstrate that the method is not even fragile against complete counterfeiting, making it incapable of detecting image tampering. Essentially, the proposed attack enables the attacker to counterfeit the watermarks belonging to any number of 4×4 blocks and embed them in arbitrary blocks. The results indicate that the proposed attack successfully deceives the method. © 2016 IEEE.
Mahdavi, M.,
Darvish morshedi hosseini, M.,
Mahdavi, M. Publication Date: 2015
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025
Performance of universal steganalysis highly depends on the features extracted from the images. Recently there have been some high-dimensional feature sets introduced in order to model a large number of dependencies between neighboring pixels and JPEG coefficients. Although using these high-dimensional models can increase detection rate, due to their dimensionality, they can induce some problems in the classification process. Furthermore, extraction of such excessively large models is time-consuming. Using a feature selection strategy can lead to selection of the most prominent features and as a result, it can decrease feature extraction time. Another advantage of feature selection can be detection of the features that should be preserved in the steganography process in order to avoid detection of steganography. In this paper, a new feature selection algorithm is suggested which utilizes two statistical measures (i.e., KS from Kolmogorov-Smirnov test and F from F-to-remove). For selecting features, the proposed method does not benefit from a classifier; therefore, it should be considered as a filter method. In the proposed method, according to F statistic which is available in F-to-remove method, a reordering is applied on the features. Afterward, the features are mutually compared using KS-test and if the distributions of the two features are equal, one of them is discarded. The comparison of the proposed method with a recently introduced filter-type method for this aim shows performance improvements in terms of the effectiveness of selected features. © 2015 IEEE.
Publication Date: 2015
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 80-86
Digital watermarking has become an important technique for copyright protection of multimedia contents. In recent years, Singular Value Decomposition (SVD) has been used as a valuable transform technique for robust digital watermarking. Designers mostly use dewatermarking tools such as Stirmark to prove the robustness of watermarking schemes. Distortions available in these tools usually degrade the quality of images and resistance against these distortions does not imply the scheme is secure. Despite of distortions, there are many types of attacks which violate security. There is a kind of attack in which the attacker needs to know the watermarking algorithm to perform the attack. In this paper, two attacks of this kind have been designed against two specific non-blind SVD-based watermarking schemes. These attacks try to subtract the added watermark or change the exact spaces where watermarks have been added. The experiments show our attacks make fewer changes to remove embedded watermarks than other distortions available in Stirmark, so the quality of attacked images is better than processed images by using Stirmark in terms of PSNR. © 2015 IEEE.
Mahdavi, M.,
Darvish morshedi hosseini, M.,
Mahdavi, M. Publication Date: 2015
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 187-194
High-dimensional feature sets proposed for steganalysis are able to model a lot of dependencies between pixels. Although these dependencies can reveal the changes induced by steganography in pixels or JPEG coefficients values, for some known steganography methods, some features may not change significantly. The significance of feature selection in steganalysis relies on two facets. First, by keeping particular features constant in steganography, a more resistant embedding scheme can be obtained; in other words, the most prominent features for detecting a particular steganography method can reveal the weaknesses of the method. Second, extraction of high-dimensional feature sets in steganalysis is a time-consuming process and this issue prevents steganalysis applicability in the real-word problems, whereas in some cases feature selection might lead to reduction of feature extraction time. In this paper, a novel and simple method is suggested in order to select the most prominent features from the feature sets. The aim of the proposed method is not to increase the classification accuracy, but it aims at decreasing the negative effect of removing weak-discriminant features on classification accuracy and therefore at decreasing the classification complexity along with minimum degradation. Also another goal is to find the features that should be preserved in steganography process in order to avoid detection of steganography. The proposed method begins with sorting the features based on a selective ranking function, and then by considering a value as threshold, only the features that can increase classification accuracy more than the threshold value are selected. Due to its similarities to Forward selection algorithm, it is compared with this selection. The comparison results showed improvement in terms of the selected features. © 2015 IEEE.
Publication Date: 2015
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 12-17
Steganography and Covert channels, their design and detection has been one of the most important issues in security systems and computer networks. One of the media that is used for steganography is JPEG images. The purpose of many steganography techniques in pictures is to insert more covert information in image's pixels with few changes. One of methods in JPEG steganography is EMD (Exploiting Modification Direction). EMD tries to increase embedding efficiency which means embed more hidden bits for each change. On the other hand, due to widespread use of wireless networks, designing covert channels in these networks has also attracted a lot of attention. Local wireless networks have a high degree of randomness in the selection of back off times in their collision avoidance algorithm. This feature can be used to create covert channels. In this article, the idea of EMD method is discussed to create a covert channel in wireless networks. The goal is to achieve high throughput with high degree of security. © 2015 IEEE.
Shahgholi, B.,
Eftekhari, S.,
Shahgholi ghahfarokhi, B.,
Moghim, N. Publication Date: 2015
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025
The tremendous growth of the Internet traffic and the rapid changes that have occurred in the way people using it to access massive contents, instruct the research community to data oriented networks. Over the last few years, various data oriented network architectures have emerged to fulfill the demand for a more scalable content distribution. Content-Centric Networking (CCN) is an architecture that has attracted a good consideration. A significant portion of Internet traffic growth relates to diverse types of multimedia applications, including live TV. In fact, delivering television services over the existing Internet protocol (IPTV) has been commercialized for more than a decade. Emigrating to the new network generation requires well analysis of the current applications. Nevertheless, CCN still suffers from the shortage of suitable simulators. We have developed a new modular, component based CCN simulator specifically optimized for live TV modeling. We have published it as an open source utility. In this paper, we present our simulator and evaluate its capabilities on the simulation of live video streaming over CCN. © 2015 IEEE.
Publication Date: 2015
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025
Digital watermarking is widely used to verify the authenticity or integrity of multimedia contents, such as images and videos. In the last few years, a well-known numerical tool called singular value decomposition (SVD) has received much attention from the watermarking community. Designers of SVD-based schemes usually use dewatermarking tools such as Stirmark to prove the robustness of their schemes. Although, these tools are valuable but their attacks usually reduce the quality of watermarked images. On the other hand, there is another group of attacks in which the attacker needs to know the watermarking algorithm to perform the attack. In this paper, an attack in this group has been designed to remove the watermark from a specific SVD-based watermarking scheme [1]. The attack tries to change the exact space where watermark has been embedded. The experiments show our attack makes fewer changes to remove embedded watermark than other distortions available in Stirmark, so the quality of attacked image is better than processed image by using Stirmark in terms of PSNR. © 2015 IEEE.
Publication Date: 2014
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 519-524
Due to the high costs of deploying and testing VANETs, simulations are required for development and evaluation of new protocols at any layer of WAVE protocol stack. Although there are powerful tools to simulate VANETs and especially the IEEE 802.11p/1609.4 DSRC/WAVE protocols, however, they are not free or not publically available. NS-2 Network Simulator is a free and widely accepted simulator used by researchers to simulate DSRC/WAVE. To the best of our knowledge, current version of NS-2 and developed simulation tools based on NS-2 do not appropriately support for the multichannel operation of the IEEE 802.11p/1609.4. In this paper, a new developed simulation tool based on NS-2 is introduced that provides a more realistic implementation of DSRC multiple channels and IEEE 1609.4 multichannel operation. The developed simulation tool has been tested through some simulation experiments. © 2014 IEEE.
Publication Date: 2014
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 1071-1076
Improving cell coverage and network capacity are main issues in LTE networks. By the emergence of heterogeneous cellular networks with different cell size, femtocells have been regarded as a low cost solution to improve poor indoor coverage for home users. However, as Femto Access Points (FAPs) are installed by users, self-organized techniques are needed for allocation of radio resources to femtocells. On the other hand, Fractional Frequency reuse (FFR) has been considered to improve spectral efficiency and quality of edge users in heterogeneous networks (HetNets). In conventional FFR methods, the macrocell area is partitioned into some regions and certain fractions of radio resources are considered for macrocell!femtocell users in each region. Therefore, radio resources are allocated to femtocell!macrocell users based on their region of presence without addressing the density of users in that region and consequently the interference level. In this paper, a new self-organized fractional resource allocation method is proposed for femtocells. The proposed method is based on Learning Automata where FAPs learn to choose the best fraction based on the feedback of femtocell users. Simulation results confirm that the proposed radio resource allocation method improves spectral efficiency and decreases the outage probability compared to conventional Strict FFR method. © 2014 IEEE.
Mahdavi, M.,
Khorramdin M.,
Amini M.,
Torabi, N.,
Mahdavi, M. Publication Date: 2014
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 961-965
Embedding an identifying data into digitized music, video or image is known as digital watermarking. Reversible image watermarking is used to embed and extract hidden data to and from the watermarked image without any distortion to the original. In this paper, we propose an improved reversible watermarking scheme using an additive interpolation technique, which increases embedding capacity with inconspicuous degradation of image quality. Unlike former watermarking schemes, a new interpolation-error is exploited to embed bit "1" or "0" using additive expansion or leaving it unchanged. Consequently, original image is remained at a high level of quality. Moreover, a cryptographic scheme is utilized to prevent an adversary from accessing embedded data. Therefore, achieving watermarked data will not be straightforward. The experimental results show that the proposed method in addition to greater embedding capacity, has higher image fidelity compared to previous schemes. © 2014 IEEE.
Publication Date: 2014
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 747-752
Wireless access in vehicular environment brings new challenges. One of the most common issues is about designing a channel access scheme for Vehicle-to-Vehicle (V2V) communications, which is affected by several number of factors such as high density of vehicles, direction of traveling, location and speed of vehicles, instability of communication links, and absence of fixed infrastructures. The IEEE 802.11p/1609.4 WAVE protocols aim to cope with these issues through adopting a contention-based approach. However, various simulation experiments vote the weakness of the IEEE 802.11p/1609.4 protocols to provide scalability, reliability, predictable delay and fairness. Meanwhile, contention-free methods have been offered by researchers to address these requirements. However, there are only a few number of contention-free schemes for VANETs in the literature, which consider fairness. In this paper, a Time Division Multiple Access (TDMA) based multichannel assignment scheme is proposed to address the fairness issue in sharing wireless medium. The proposed method is an extension of TC-MAC and apart from improving fairness; it tries to preserve the bandwidth utilization via an enhanced slot reservation mechanism. © 2014 IEEE.
Publication Date: 2021
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 358-362
In terms of wireless networks and energy consumption, which is the main basis of Internet of Things (IoT) devices, low-power networks (LPWANs) are considered as a suitable solution for IoT applications. The most important LPWAN protocols are SigFox, NB-IoT(NarrowBand-IoT), and LoRa. LoRa is more popular than others because of industry supports such as LoRa Alliance, IBM, Cisco, etc. The large number of devices that communicate with each other in IoT applications has raised concerns about resources availability and the suitable technologies for managing these resources in LoRaWAN networks. There are many applications in the IoT in which the time to receive information from devices (sensors) can be adjusted based on the context information and the internal state of the system in a way that reduces collisions. Therefore, the time of sending information by sensors can be adjusted to some extent, so that sensors have less collisions during transmissions, and thus the problem of scalability of the system is almost solved. In this research, we have presented a framework to improve the performance of LoRaWAN network, in which the devices are scheduled according to their QoS requirements, the density of the network, and the context information. The performance of the proposed model is evaluated by simulations in which the sensors send packets to the server based on the proposed scheduling. Evaluations show that the proposed method has reduced congestion by 51% and energy consumption by 52% on average compared to the baseline solution. © 2021 IEEE.
Mahdavi, M.,
Khalilidan S.,
Mahdavi, M.,
Balouchestani, A.,
Moti Z.,
Hallaj, Y. Publication Date: 2020
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 229-233
Recent advents of the internet have made accessibility of people to digital data such as audio, images, and videos much easier. Meanwhile, one of the cases that adversaries take advantage of is the people's face images that are available across the web. Digital watermarking is used to authenticate the original owner of the images and protect their copyright. With the help of digital watermarking, hidden data is embedded inside the image. Recently, neural networks such as autoencoders are one of the most popular models that are used in many fields. Neural networks are capable of understanding all kinds of raw data such as images and videos. In this paper, we present a method for embedding the user's national ID in their face images using autoencoders. The proposed autoencoder is trained with a dataset contains face images. The image is coded into some code using the autoencoders' encoder. Then, the national ID is embedded in this code and the modified code is reconstructed using the decoder to form the watermarked image. To extract the watermark, the watermarked image is encoded with the encoder and the watermark is extracted. Experiment results show that our model recovers the watermark with high accuracy and it is resistant against JPEG attacks. Moreover, the quality of the watermarked images is acceptable, and their SSIM compare to the original image is about 90%. © 2020 IEEE.
Mahdavi, M.,
Balouchestani, A.,
Mahdavi, M.,
Hallaj, Y.,
Javdani D. Publication Date: 2019
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 139-143
Millions of news are being exchanged daily among people. With the appearance of the Internet, the way of broadcasting news has changed and become faster, however it caused many problems. For instance, the increase in the speed of broadcasting news leads to an increase in the speed of fake news creation. Fake news can have a huge impression on societies. Additionally, the existence of a central entity, such as news agencies, could lead to fraud in the news broadcasting process, e.g. generating fake news and publishing them for their benefits. Since Blockchain technology provides a reliable decentralized network, it can be used to publish news. In addition, Blockchain with the help of decentralized applications and smart contracts can provide a platform in which fake news can be detected through public participation. In this paper, we proposed a new method for sharing and analyzing news to detect fake news using Blockchain, called SANUB. SANUB provides features such as publishing news anonymously, news evaluation, reporter validation, fake news detection and proof of news ownership. The results of our analysis show that SANUB outperformed the existing methods. © 2019 IEEE.
Publication Date: 2019
2025 29th International Computer Conference, Computer Society of Iran, CSICC 2025pp. 1289-1293
Special characteristics of Machine-Type Communications (MTC) result in new challenges in 5G mobile networks including Medium Access Control (MAC) mechanisms. The traditional methods for contention-based MAC are unable to manage a massive number of simultaneous requests. On another side, recent methods for MTCs have not adequately attended to energy consumption of machines along with QoS. In this paper, a method for contention-based massive access protocol is presented based on p-persistent CSMA which considers both energy and delay requirement of machines, a multi-objective optimization problem is solved by BS at the beginning of each time frame in order to find the optimum value of transmission probability based on the number of machines and their distances from the BS. The optimization problem aims at achieving maximum energy efficiency and minimum access delay. To avoid the overhead of solving the optimization problem, the value of probability is temporarily updated by the machine at the end of each time slot, based on its delay requirement and the number of its recent access failures. The simulation results show a considerable increase in the number of successful transmissions, a decrease in access time and improvements in throughput and energy consumption. © 2019 IEEE.
Mahdavi, E.,
Fanian, A.,
Mirzaei, A.,
Taghiyarrenani, Zahra Publication Date: 2022
Knowledge-Based Systems (0950-7051)
Utilizing machine learning methods to detect intrusion into computer networks is a trending topic in information security research. The limitation of labeled samples is one of the challenges in this area. This challenge makes it difficult to build accurate learning models for intrusion detection. Transfer learning is one of the methods to counter such a challenge in machine learning topics. On the other hand, the emergence of new technologies and applications might bring new vulnerabilities to computer networks. Therefore, the learning process cannot occur all at once. Incremental learning is a practical standpoint to confront this challenge. This research presents a new framework for intrusion detection systems called ITL-IDS that can potentially start learning in a network without prior knowledge. It begins with an incremental clustering algorithm to detect clusters’ numbers and shape without prior assumptions about the attacks. The outcomes are candidates to transfer knowledge between other instances of ITL-IDS. In each iteration, transfer learning provides target environments with incremental knowledge. Our evaluation shows that this method can combine incremental and transfer learning to identify new attacks. © 2022
Publication Date: 2020
Computers and Security (01674048)
Alert Correlation is the process of analyzing alerts to reduce their number, eliminate false positives, detect the scenarios behind them and generate a higher perspective of the incidents. Making this process online will upgrade the classic role of alert correlation from being a post-process step to a key part of intrusion detection systems. In this article, we propose a novel two-phase model called a Real-time Alert Correlation method based on Code-books (RACC) for intrusion detection systems. First, in the offline phase, RACC pre-processes a knowledge base to propose some matrices as the main data structure of the method that we call them code-books. Instead of keeping alerts in the memory, those matrices just hold keys to the corresponding meta-alerts. An index that is based upon red-black trees is used to access matrix elements. Generating the matrices and mentioned index are independent from the alerts, so utilizing them can facilitate the alert correlation process in an online manner in phase two of the proposed model. The experiments show that compared to similar methods, RACC can significantly reduce the alert correlation time and can enable real-time alert correlation. © 2019 Elsevier Ltd
In the past decades, machine learning based intrusion detection systems have been developed. This paper discloses a new aspect of machine learning based intrusion detection systems. The proposed method detects normal and anomaly behaviors in the desired network where there are not any labeled samples as training dataset. That is while a plenty of labeled samples may exist in another network that is different from the desired network. Because of the difference between two networks, their samples produce in different manners. So, direct utilizing of labeled samples of a different network as training samples does not provide acceptable accuracy to detect anomaly behaviors in the desired network. In this paper, we propose a transfer learning based intrusion detection method which transfers knowledge between the networks and eliminates the problem of providing training samples that is a costly procedure. Comparing the experimental results with the results of a basic machine learning method (SVM) and also baseline method(DAMA) shows the effectiveness of the proposed method for transferring knowledge for intrusion detection systems. © 2018 IEEE.
Publication Date: 2015
International Journal of Data Mining and Bioinformatics (17485673)(2)
MicroRNAs (miRNAs) are a class of short RNA molecules that regulate gene expression by binding directly to messenger RNAs. Conventional approaches to miRNA target prediction estimate the accessibility of target sites and the strength of the binding miRNA by finding optimums of some energy models, which involves O(n3) computations. Alternatively, we narrow down potential binding sites of miRNAs to suboptimal hits of a pairwise alignment algorithm called Fitting Alignment in O(n2). We invoke a same algorithm, once for all candidate sites to measure the site accessibilities. These features are applied to a binary classifier being learned to predict true associations between miRNAs and target genes. Training the classifier requires the negative samples indicating non-affected genes. The experiments verifying such negative associations have been rarely performed, so we exploit tissue-specific gene expression data to impute the negative associations. The recall rate of our method is above 70% (at precision 85%). Copyright © 2015 Inderscience Enterprises Ltd.
Publication Date: 2024
Information (Switzerland) (20782489)15(8)
During automated negotiations, intelligent software agents act based on the preferences of their proprietors, interdicting direct preference exposure. The agent can be armed with a component of an opponent’s modeling features to reduce the uncertainty in the negotiation, but how negotiating agents with a single-peaked preference direct our attention has not been considered. Here, we first investigate the proper representation of single-peaked preferences and implementation of single-peaked agents within bidder agents using different instances of general single-peaked functions. We evaluate the modeling of single-peaked preferences and bidders in automated negotiating agents. Through experiments, we reveal that most of the opponent models can model our benchmark single-peaked agents with similar efficiencies. However, the accuracies differ among the models and in different rival batches. The perceptron-based P1 model obtained the highest accuracy, and the frequency-based model Randomdance outperformed the other competitors in most other performance measures. © 2024 by the authors.
Feedback is an essential component of learning, as it helps students identify their strengths and weaknesses, and improve their performance. However, the students may not be able to understand how their work has been judged. One way to address this issue is to let the students assess and comment on the work of their peers, Peer Assessment (PA). PA has benefits such as enhancing learning outcomes, developing self-assess and critical thinking skills, and fostering collaboration. However, PA also poses some challenges such as ensuring fairness, anonymity, and reliability. In this study, we designed and implemented an anonymous electronic PA for a few classes participating EU-Iran STEM/UNITEL project. We used several digital tools to simulate a double-blind PA process, and a rubric based on the students' own criteria and weights. The grades collected from 70 students represent positive feedback. We discuss the functionality, advantages, and limitations of our approach. © 2024 IEEE.
