Rivista e Edizione

Volume 22 (2022): Edizione 3 (September 2022)

Volume 22 (2022): Edizione 2 (June 2022)

Volume 22 (2022): Edizione 1 (March 2022)

Volume 21 (2021): Edizione 4 (December 2021)

Volume 21 (2021): Edizione 3 (September 2021)

Volume 21 (2021): Edizione 2 (June 2021)

Volume 21 (2021): Edizione 1 (March 2021)

Volume 20 (2020): Edizione 6 (December 2020)
Special Edizione on New Developments in Scalable Computing

Volume 20 (2020): Edizione 5 (December 2020)
Special issue on Innovations in Intelligent Systems and Applications

Volume 20 (2020): Edizione 4 (November 2020)

Volume 20 (2020): Edizione 3 (September 2020)

Volume 20 (2020): Edizione 2 (June 2020)

Volume 20 (2020): Edizione 1 (March 2020)

Volume 19 (2019): Edizione 4 (November 2019)

Volume 19 (2019): Edizione 3 (September 2019)

Volume 19 (2019): Edizione 2 (June 2019)

Volume 19 (2019): Edizione 1 (March 2019)

Volume 18 (2018): Edizione 5 (May 2018)
Special Thematic Edizione on Optimal Codes and Related Topics

Volume 18 (2018): Edizione 4 (November 2018)

Volume 18 (2018): Edizione 3 (September 2018)

Volume 18 (2018): Edizione 2 (June 2018)

Volume 18 (2018): Edizione 1 (March 2018)

Volume 17 (2017): Edizione 5 (December 2017)
Special Edizione With Selected Papers From The Workshop “Two Years Avitohol: Advanced High Performance Computing Applications 2017

Volume 17 (2017): Edizione 4 (November 2017)

Volume 17 (2017): Edizione 3 (September 2017)

Volume 17 (2017): Edizione 2 (June 2017)

Volume 17 (2017): Edizione 1 (March 2017)

Volume 16 (2016): Edizione 6 (December 2016)
Special issue with selection of extended papers from 6th International Conference on Logistic, Informatics and Service Science LISS’2016

Volume 16 (2016): Edizione 5 (October 2016)
Edizione Title: Special Edizione on Application of Advanced Computing and Simulation in Information Systems

Volume 16 (2016): Edizione 4 (December 2016)

Volume 16 (2016): Edizione 3 (September 2016)

Volume 16 (2016): Edizione 2 (June 2016)

Volume 16 (2016): Edizione 1 (March 2016)

Volume 15 (2015): Edizione 7 (December 2015)
Special Edizione on Information Fusion

Volume 15 (2015): Edizione 6 (December 2015)
Special Edizione on Logistics, Informatics and Service Science

Volume 15 (2015): Edizione 5 (April 2015)
Special Edizione on Control in Transportation Systems

Volume 15 (2015): Edizione 4 (November 2015)

Volume 15 (2015): Edizione 3 (September 2015)

Volume 15 (2015): Edizione 2 (June 2015)

Volume 15 (2015): Edizione 1 (March 2015)

Volume 14 (2014): Edizione 5 (December 2014)
Special Edizione

Volume 14 (2014): Edizione 4 (December 2014)

Volume 14 (2014): Edizione 3 (September 2014)

Volume 14 (2014): Edizione 2 (June 2014)

Volume 14 (2014): Edizione 1 (March 2014)

Volume 13 (2013): Edizione Special-Edizione (December 2013)

Volume 13 (2013): Edizione 4 (December 2013)
The publishing of the present issue (Volume 13, No 4, 2013) of the journal “Cybernetics and Information Technologies” is financially supported by FP7 project “Advanced Computing for Innovation” (ACOMIN), grant agreement 316087 of Call FP7 REGPOT-2012-2013-1.

Volume 13 (2013): Edizione 3 (September 2013)

Volume 13 (2013): Edizione 2 (June 2013)

Volume 13 (2013): Edizione 1 (March 2013)

Volume 12 (2012): Edizione 4 (December 2012)

Volume 12 (2012): Edizione 3 (September 2012)

Volume 12 (2012): Edizione 2 (June 2012)

Volume 12 (2012): Edizione 1 (March 2012)

Dettagli della rivista
Formato
Rivista
eISSN
1314-4081
Pubblicato per la prima volta
13 Mar 2012
Periodo di pubblicazione
4 volte all'anno
Lingue
Inglese

Cerca

Volume 22 (2022): Edizione 1 (March 2022)

Dettagli della rivista
Formato
Rivista
eISSN
1314-4081
Pubblicato per la prima volta
13 Mar 2012
Periodo di pubblicazione
4 volte all'anno
Lingue
Inglese

Cerca

12 Articoli
Accesso libero

Hiding Sensitive High Utility and Frequent Itemsets Based on Constrained Intersection Lattice

Pubblicato online: 10 Apr 2022
Pagine: 3 - 23

Astratto

Abstract

Hiding high utility and frequent itemset is the method used to preserve sensitive knowledge from being revealed by pattern mining process. Its goal is to remove sensitive high utility and frequent itemsets from a database before sharing it for data mining purposes while minimizing the side effects. The current methods succeed in the hiding goal but they cause high side effects. This paper proposes a novel algorithm, named HSUFIBL, that applies a heuristic for finding victim item based on the constrained intersection lattice theory. This algorithm specifies exactly the condition that allows the application of utility reduction or support reduction method, the victim item, and the victim transaction for the hiding process so that the process needs the fewest data modifications and gives the lowest number of lost non-sensitive itemsets. The experimental results indicate that the HSUFIBL algorithm achieves better performance than previous works in minimizing the side effect.

Parole chiave

  • High utility mining
  • High utility and frequent itemset
  • Sensitive high utility and frequent itemset hiding
  • Privacy-preserving utility mining
Accesso libero

Deterministic Centroid Localization for Improving Energy Efficiency in Wireless Sensor Networks

Pubblicato online: 10 Apr 2022
Pagine: 24 - 39

Astratto

Abstract

Wireless sensor networks are an enthralling field of study with numerous applications. A Wireless Sensor Network (WSN) is used to monitor real-time scenarios such as weather, temperature, humidity, and military surveillance. A WSN is composed of several sensor nodes that are responsible for sensing, aggregating, and transmitting data in the system, in which it has been deployed. These sensors are powered by small batteries because they are small. Managing power consumption and extending network life is a common challenge in WSNs. Data transmission is a critical process in a WSN that consumes the majority of the network’s resources. Since the cluster heads in the network are in charge of data transmission, they require more energy. We need to know where these CHs are deployed in order to calculate how much energy they use. The deployment of a WSN can be either static or random. Although most researchers focus on random deployment, this paper applies the proposed Deterministic Centroid algorithm for static deployment. Based on the coverage of the deployment area, this algorithm places the sensors in a predetermined location. The simulation results show how this algorithm generates balanced clusters, improves coverage, and saves energy.

Parole chiave

  • Sensors
  • energy
  • centroid
  • DV Hop
  • WCL
  • deployment
  • clustering
Accesso libero

A Proposal for Honeyword Generation via Meerkat Clan Algorithm

Pubblicato online: 10 Apr 2022
Pagine: 40 - 59

Astratto

Abstract

An effective password cracking detection system is the honeyword system. The Honeyword method attempts to increase the security of hashed passwords by making password cracking easier to detect. Each user in the system has many honeywords in the password database. If the attacker logs in using a honeyword, a quiet alert trigger indicates that the password database has been hacked. Many honeyword generation methods have been proposed, they have a weakness in generating process, do not support all honeyword properties, and have many honeyword issues. This article proposes a novel method to generate honeyword using the meerkat clan intelligence algorithm, a metaheuristic swarm intelligence algorithm. The proposed generation methods will improve the honeyword generating process, enhance the honeyword properties, and solve the issues of previous methods. This work will show some previous generation methods, explain the proposed method, discuss the experimental results and compare the new one with the prior ones.

Parole chiave

  • Honeyword
  • password
  • metaheuristic
  • swarm
  • meerkat
Accesso libero

Enhancеd Analysis Approach to Detect Phishing Attacks During COVID-19 Crisis

Pubblicato online: 10 Apr 2022
Pagine: 60 - 76

Astratto

Abstract

Public health responses to the COVID-19 pandemic since March 2020 have led to lockdowns and social distancing in most countries around the world, with a shift from the traditional work environment to virtual one. Employees have been encouraged to work from home where possible to slow down the viral infection. The massive increase in the volume of professional activities executed online has posed a new context for cybercrime, with the increase in the number of emails and phishing websites. Phishing attacks have been broadened and extended through years of pandemics COVID-19. This paper presents a novel approach for detecting phishing Uniform Resource Locators (URLs) applying the Gated Recurrent Unit (GRU), a fast and highly accurate phishing classifier system. Comparative analysis of the GRU classification system indicates better accuracy (98.30%) than other classifier systems.

Parole chiave

  • Cybersecurity
  • COVID-19
  • phishing attack
  • cybercrime
Accesso libero

Data Fusion and the Impact of Group Mobility on Load Distribution on MRHOF and OF0

Pubblicato online: 10 Apr 2022
Pagine: 77 - 94

Astratto

Abstract

Many routing algorithms proposed for IoT are based on modifications on RPL objective functions and trickle algorithms. However, there is a lack of an in-depth study to examine the impact of mobility on routing protocols based on MRHOF and OF0 algorithms. This paper examines the impact of group mobility on these algorithms, also examines their ability in distributing the load and the impact of varying traffic with the aid of simulations using the well-known Cooja simulator. The two algorithms exhibit similar performance for various metrics for low traffic rates and low mobility speed. However, when the traffic rate becomes relatively high, OF0 performance merits appear, in terms of throughput, packet load deviation, power deviation, and CPU power deviation. The mobility with higher speeds helps MRHOF to enhance its throughput and load deviation. The mobility allowed MRHOF to demonstrate better packets load deviation.

Parole chiave

  • IoT
  • MRHOF
  • OF0
  • load distribution
Accesso libero

Citation and Similarity in Academic Texts: Colombian Engineering Case

Pubblicato online: 10 Apr 2022
Pagine: 95 - 103

Astratto

Abstract

This article provides the results of a citation determinants model for a set of academic engineering texts from Colombia. The model establishes the determinants of the probability that a text receives at least one citation through the relationship among previous citations, journal characteristics, the author and the text. Through a similarity matrix constructed by Latent Semantic Analysis (LSA), a similarity variable has been constructed to capture the fact that the texts have similar titles, abstracts and keywords to the most cited texts. The results show: i) joint significance of the variables selected to characterize the text; ii) direct relationship of the citation with similarity of keywords, published in an IEEE journal, research article, more than one author; and authored by at least one foreign author; and iii) inverse relationship between the probability of citation with the similarity of abstracts, published in 2016 or 2017, and published in a Colombian journal.

Parole chiave

  • Latent semantic analysis
  • text similarity
  • citation determinants
  • bibliometrics
Accesso libero

Combination of Resnet and Spatial Pyramid Pooling for Musical Instrument Identification

Pubblicato online: 10 Apr 2022
Pagine: 104 - 116

Astratto

Abstract

Identifying similar objects is one of the most challenging tasks in computer vision image recognition. The following musical instruments will be recognized in this study: French horn, harp, recorder, bassoon, cello, clarinet, erhu, guitar saxophone, trumpet, and violin. Numerous musical instruments are identical in size, form, and sound. Further, our works combine Resnet 50 with Spatial Pyramid Pooling (SPP) to identify musical instruments that are similar to one another. Next, the Resnet 50 and Resnet 50 SPP model evaluation performance includes the Floating-Point Operations (FLOPS), detection time, mAP, and IoU. Our work can increase the detection performance of musical instruments similar to one another. The method we propose, Resnet 50 SPP, shows the highest average accuracy of 84.64% compared to the results of previous studies.

Parole chiave

  • Resnet 50
  • Resnet 50 SPP
  • spatial pyramid pooling
  • musical instruments
  • similar object
Accesso libero

Early Student-at-Risk Detection by Current Learning Performance and Learning Behavior Indicators

Pubblicato online: 10 Apr 2022
Pagine: 117 - 133

Astratto

Abstract

The article is focused on the problem of early prediction of students’ learning failures with the purpose of their possible prevention by timely introducing supportive measures. We propose an approach to designing a predictive model for an academic course or module taught in a blended learning format. We introduce certain requirements to predictive models concerning their applicability to the educational process such as interpretability, actionability, and adaptability to a course design. We test three types of classifiers meeting these requirements and choose the one that provides best performance starting from the early stages of the semester, and therefore provides various opportunities to timely support at-risk students. Our empirical studies confirm that the proposed approach is promising for the development of an early warning system in a higher education institution. Such systems can positively influence student retention rates and enhance learning and teaching experience for a long term.

Parole chiave

  • Learning analytics
  • Learning success
  • Learning failure
  • Student-at-risk
  • Early warning system
  • Bayesian network
  • k-Nearest Neighbors
  • Linear discriminant analysis
Accesso libero

Hy-MOM: Hybrid Recommender System Framework Using Memory-Based and Model-Based Collaborative Filtering Framework

Pubblicato online: 10 Apr 2022
Pagine: 134 - 150

Astratto

Abstract

Lack of personalization, rating sparsity, and cold start are commonly seen in e-Learning based recommender systems. The proposed work here suggests a personalized fused recommendation framework for e-Learning. The framework consists of a two-fold approach to generate recommendations. Firstly, it attempts to find the neighbourhood of similar learners based on certain learner characteristics by applying a user-based collaborative filtering approach. Secondly, it generates a matrix of ratings given by the learners. The outcome of the first stage is merged with the second stage to generate recommendations for the learner. Learner characteristics, namely knowledge level, learning style, and learner preference, have been considered to bring in the personalization factor on the recommendations. As the stochastic gradient approach predicts the learner-course rating matrix, it helps overcome the rating sparsity and cold-start issues. The fused model is compared with traditional stand-alone methods and shows performance improvement.

Parole chiave

  • Recommender systems
  • e-Learning
  • Personalized
  • Fused model
  • Stochastic Gradient Descent
Accesso libero

Blockchain-Enabled Supply-Chain in Crop Production Framework

Pubblicato online: 10 Apr 2022
Pagine: 151 - 170

Astratto

Abstract

The purpose of this paper is to propose an approach to blockchain-enabled supply-chain model for a smart crop production framework. The defined tasks are: (1) analysis of blockchain ecosystem as a network of stakeholders and as an infrastructure of technical and logical elements; (2) definition of a supply-chain model; (3) design of blockchain reference infrastructure; (4) description of blockchain information channels with smart contracts basic functionalities. The results presented include: а supply-chain model facilitating seeds certification process, monitoring and supervision of the grain process, provenance and as optional interactions with regulatory bodies, logistics and financial services; the three level blockchain reference infrastructure and a blockchain-enabled supply-chain supporting five information channels with nine participants and smart contracts. An account management user application tool, the general descriptions of smart contract basic functionalities and a selected parts of one smart contract code are provided as examples.

Parole chiave

  • Blockchains
  • blockchain-enabled supply-chain
  • smart contracts
  • smart crop production
  • EOSIO platform
Accesso libero

Long Short Term Memory Neural Network-Based Model Construction and Fne-Tuning for Air Quality Parameters Prediction

Pubblicato online: 10 Apr 2022
Pagine: 171 - 189

Astratto

Abstract

Air pollution has increased worries regarding health and ecosystems. Precise prediction of air quality parameters can assist in the effective action of air pollution control and prevention. In this work, a deep learning framework is proposed to predict parameters such as fine particulate matter and carbon monoxide. Long Short Term Memory (LSTM) neural network-based model that processes sequences in forward and backward direction to consider the influence of timesteps in both directions is employed. For further learning, unidirectional layers’ stacking is implemented. The performance of the model is optimized by fine-tuning hyperparameters, regularization techniques for overfitting resolution, and various merging options for the bidirectional input layer. The proposed model achieves good optimization and performs better than the simple LSTM and a Recurrent Neural Network (RNN) based model. Moreover, an attention-based mechanism is adopted to focus on more significant timesteps for prediction. The self-attention approach improves performance further and works well especially for longer sequences and extended time horizons. Experiments are conducted using real-world data collected, and results are evaluated using the mean square error loss function.

Parole chiave

  • Air quality forecasting
  • Air pollution forecasting
  • Deep learning
  • Long short term memory
  • attention
Accesso libero

ESAR, An Expert Shoplifting Activity Recognition System

Pubblicato online: 10 Apr 2022
Pagine: 190 - 200

Astratto

Abstract

Shoplifting is a troubling and pervasive aspect of consumers, causing great losses to retailers. It is the theft of goods from the stores/shops, usually by hiding the store item either in the pocket or in carrier bag and leaving without any payment. Revenue loss is the most direct financial effect of shoplifting. Therefore, this article introduces an Expert Shoplifting Activity Recognition (ESAR) system to reduce shoplifting incidents in stores/shops. The system being proposed seamlessly examines each frame in video footage and alerts security personnel when shoplifting occurs. It uses dual-stream convolutional neural network to extract appearance and salient motion features in the video sequences. Here, optical flow and gradient components are used to extract salient motion features related to shoplifting movement in the video sequence. Long Short Term Memory (LSTM) based deep learner is modeled to learn the extracted features in the time domain for distinguishing person actions (i.e., normal and shoplifting). Analyzing the model behavior for diverse modeling environments is an added contribution of this paper. A synthesized shoplifting dataset is used here for experimentations. The experimental outcomes show that the proposed approach attains better consequences up to 90.26% detection accuracy compared to the other prevalent approaches.

Parole chiave

  • Automated surveillance system
  • Human Activity Recognition (HAR)
  • Histogram of Oriented Gradient (HOG)
  • Optical Flow
  • Convolutional Neural Network (CNN)
  • Long Short Term Memory (LSTM)
12 Articoli
Accesso libero

Hiding Sensitive High Utility and Frequent Itemsets Based on Constrained Intersection Lattice

Pubblicato online: 10 Apr 2022
Pagine: 3 - 23

Astratto

Abstract

Hiding high utility and frequent itemset is the method used to preserve sensitive knowledge from being revealed by pattern mining process. Its goal is to remove sensitive high utility and frequent itemsets from a database before sharing it for data mining purposes while minimizing the side effects. The current methods succeed in the hiding goal but they cause high side effects. This paper proposes a novel algorithm, named HSUFIBL, that applies a heuristic for finding victim item based on the constrained intersection lattice theory. This algorithm specifies exactly the condition that allows the application of utility reduction or support reduction method, the victim item, and the victim transaction for the hiding process so that the process needs the fewest data modifications and gives the lowest number of lost non-sensitive itemsets. The experimental results indicate that the HSUFIBL algorithm achieves better performance than previous works in minimizing the side effect.

Parole chiave

  • High utility mining
  • High utility and frequent itemset
  • Sensitive high utility and frequent itemset hiding
  • Privacy-preserving utility mining
Accesso libero

Deterministic Centroid Localization for Improving Energy Efficiency in Wireless Sensor Networks

Pubblicato online: 10 Apr 2022
Pagine: 24 - 39

Astratto

Abstract

Wireless sensor networks are an enthralling field of study with numerous applications. A Wireless Sensor Network (WSN) is used to monitor real-time scenarios such as weather, temperature, humidity, and military surveillance. A WSN is composed of several sensor nodes that are responsible for sensing, aggregating, and transmitting data in the system, in which it has been deployed. These sensors are powered by small batteries because they are small. Managing power consumption and extending network life is a common challenge in WSNs. Data transmission is a critical process in a WSN that consumes the majority of the network’s resources. Since the cluster heads in the network are in charge of data transmission, they require more energy. We need to know where these CHs are deployed in order to calculate how much energy they use. The deployment of a WSN can be either static or random. Although most researchers focus on random deployment, this paper applies the proposed Deterministic Centroid algorithm for static deployment. Based on the coverage of the deployment area, this algorithm places the sensors in a predetermined location. The simulation results show how this algorithm generates balanced clusters, improves coverage, and saves energy.

Parole chiave

  • Sensors
  • energy
  • centroid
  • DV Hop
  • WCL
  • deployment
  • clustering
Accesso libero

A Proposal for Honeyword Generation via Meerkat Clan Algorithm

Pubblicato online: 10 Apr 2022
Pagine: 40 - 59

Astratto

Abstract

An effective password cracking detection system is the honeyword system. The Honeyword method attempts to increase the security of hashed passwords by making password cracking easier to detect. Each user in the system has many honeywords in the password database. If the attacker logs in using a honeyword, a quiet alert trigger indicates that the password database has been hacked. Many honeyword generation methods have been proposed, they have a weakness in generating process, do not support all honeyword properties, and have many honeyword issues. This article proposes a novel method to generate honeyword using the meerkat clan intelligence algorithm, a metaheuristic swarm intelligence algorithm. The proposed generation methods will improve the honeyword generating process, enhance the honeyword properties, and solve the issues of previous methods. This work will show some previous generation methods, explain the proposed method, discuss the experimental results and compare the new one with the prior ones.

Parole chiave

  • Honeyword
  • password
  • metaheuristic
  • swarm
  • meerkat
Accesso libero

Enhancеd Analysis Approach to Detect Phishing Attacks During COVID-19 Crisis

Pubblicato online: 10 Apr 2022
Pagine: 60 - 76

Astratto

Abstract

Public health responses to the COVID-19 pandemic since March 2020 have led to lockdowns and social distancing in most countries around the world, with a shift from the traditional work environment to virtual one. Employees have been encouraged to work from home where possible to slow down the viral infection. The massive increase in the volume of professional activities executed online has posed a new context for cybercrime, with the increase in the number of emails and phishing websites. Phishing attacks have been broadened and extended through years of pandemics COVID-19. This paper presents a novel approach for detecting phishing Uniform Resource Locators (URLs) applying the Gated Recurrent Unit (GRU), a fast and highly accurate phishing classifier system. Comparative analysis of the GRU classification system indicates better accuracy (98.30%) than other classifier systems.

Parole chiave

  • Cybersecurity
  • COVID-19
  • phishing attack
  • cybercrime
Accesso libero

Data Fusion and the Impact of Group Mobility on Load Distribution on MRHOF and OF0

Pubblicato online: 10 Apr 2022
Pagine: 77 - 94

Astratto

Abstract

Many routing algorithms proposed for IoT are based on modifications on RPL objective functions and trickle algorithms. However, there is a lack of an in-depth study to examine the impact of mobility on routing protocols based on MRHOF and OF0 algorithms. This paper examines the impact of group mobility on these algorithms, also examines their ability in distributing the load and the impact of varying traffic with the aid of simulations using the well-known Cooja simulator. The two algorithms exhibit similar performance for various metrics for low traffic rates and low mobility speed. However, when the traffic rate becomes relatively high, OF0 performance merits appear, in terms of throughput, packet load deviation, power deviation, and CPU power deviation. The mobility with higher speeds helps MRHOF to enhance its throughput and load deviation. The mobility allowed MRHOF to demonstrate better packets load deviation.

Parole chiave

  • IoT
  • MRHOF
  • OF0
  • load distribution
Accesso libero

Citation and Similarity in Academic Texts: Colombian Engineering Case

Pubblicato online: 10 Apr 2022
Pagine: 95 - 103

Astratto

Abstract

This article provides the results of a citation determinants model for a set of academic engineering texts from Colombia. The model establishes the determinants of the probability that a text receives at least one citation through the relationship among previous citations, journal characteristics, the author and the text. Through a similarity matrix constructed by Latent Semantic Analysis (LSA), a similarity variable has been constructed to capture the fact that the texts have similar titles, abstracts and keywords to the most cited texts. The results show: i) joint significance of the variables selected to characterize the text; ii) direct relationship of the citation with similarity of keywords, published in an IEEE journal, research article, more than one author; and authored by at least one foreign author; and iii) inverse relationship between the probability of citation with the similarity of abstracts, published in 2016 or 2017, and published in a Colombian journal.

Parole chiave

  • Latent semantic analysis
  • text similarity
  • citation determinants
  • bibliometrics
Accesso libero

Combination of Resnet and Spatial Pyramid Pooling for Musical Instrument Identification

Pubblicato online: 10 Apr 2022
Pagine: 104 - 116

Astratto

Abstract

Identifying similar objects is one of the most challenging tasks in computer vision image recognition. The following musical instruments will be recognized in this study: French horn, harp, recorder, bassoon, cello, clarinet, erhu, guitar saxophone, trumpet, and violin. Numerous musical instruments are identical in size, form, and sound. Further, our works combine Resnet 50 with Spatial Pyramid Pooling (SPP) to identify musical instruments that are similar to one another. Next, the Resnet 50 and Resnet 50 SPP model evaluation performance includes the Floating-Point Operations (FLOPS), detection time, mAP, and IoU. Our work can increase the detection performance of musical instruments similar to one another. The method we propose, Resnet 50 SPP, shows the highest average accuracy of 84.64% compared to the results of previous studies.

Parole chiave

  • Resnet 50
  • Resnet 50 SPP
  • spatial pyramid pooling
  • musical instruments
  • similar object
Accesso libero

Early Student-at-Risk Detection by Current Learning Performance and Learning Behavior Indicators

Pubblicato online: 10 Apr 2022
Pagine: 117 - 133

Astratto

Abstract

The article is focused on the problem of early prediction of students’ learning failures with the purpose of their possible prevention by timely introducing supportive measures. We propose an approach to designing a predictive model for an academic course or module taught in a blended learning format. We introduce certain requirements to predictive models concerning their applicability to the educational process such as interpretability, actionability, and adaptability to a course design. We test three types of classifiers meeting these requirements and choose the one that provides best performance starting from the early stages of the semester, and therefore provides various opportunities to timely support at-risk students. Our empirical studies confirm that the proposed approach is promising for the development of an early warning system in a higher education institution. Such systems can positively influence student retention rates and enhance learning and teaching experience for a long term.

Parole chiave

  • Learning analytics
  • Learning success
  • Learning failure
  • Student-at-risk
  • Early warning system
  • Bayesian network
  • k-Nearest Neighbors
  • Linear discriminant analysis
Accesso libero

Hy-MOM: Hybrid Recommender System Framework Using Memory-Based and Model-Based Collaborative Filtering Framework

Pubblicato online: 10 Apr 2022
Pagine: 134 - 150

Astratto

Abstract

Lack of personalization, rating sparsity, and cold start are commonly seen in e-Learning based recommender systems. The proposed work here suggests a personalized fused recommendation framework for e-Learning. The framework consists of a two-fold approach to generate recommendations. Firstly, it attempts to find the neighbourhood of similar learners based on certain learner characteristics by applying a user-based collaborative filtering approach. Secondly, it generates a matrix of ratings given by the learners. The outcome of the first stage is merged with the second stage to generate recommendations for the learner. Learner characteristics, namely knowledge level, learning style, and learner preference, have been considered to bring in the personalization factor on the recommendations. As the stochastic gradient approach predicts the learner-course rating matrix, it helps overcome the rating sparsity and cold-start issues. The fused model is compared with traditional stand-alone methods and shows performance improvement.

Parole chiave

  • Recommender systems
  • e-Learning
  • Personalized
  • Fused model
  • Stochastic Gradient Descent
Accesso libero

Blockchain-Enabled Supply-Chain in Crop Production Framework

Pubblicato online: 10 Apr 2022
Pagine: 151 - 170

Astratto

Abstract

The purpose of this paper is to propose an approach to blockchain-enabled supply-chain model for a smart crop production framework. The defined tasks are: (1) analysis of blockchain ecosystem as a network of stakeholders and as an infrastructure of technical and logical elements; (2) definition of a supply-chain model; (3) design of blockchain reference infrastructure; (4) description of blockchain information channels with smart contracts basic functionalities. The results presented include: а supply-chain model facilitating seeds certification process, monitoring and supervision of the grain process, provenance and as optional interactions with regulatory bodies, logistics and financial services; the three level blockchain reference infrastructure and a blockchain-enabled supply-chain supporting five information channels with nine participants and smart contracts. An account management user application tool, the general descriptions of smart contract basic functionalities and a selected parts of one smart contract code are provided as examples.

Parole chiave

  • Blockchains
  • blockchain-enabled supply-chain
  • smart contracts
  • smart crop production
  • EOSIO platform
Accesso libero

Long Short Term Memory Neural Network-Based Model Construction and Fne-Tuning for Air Quality Parameters Prediction

Pubblicato online: 10 Apr 2022
Pagine: 171 - 189

Astratto

Abstract

Air pollution has increased worries regarding health and ecosystems. Precise prediction of air quality parameters can assist in the effective action of air pollution control and prevention. In this work, a deep learning framework is proposed to predict parameters such as fine particulate matter and carbon monoxide. Long Short Term Memory (LSTM) neural network-based model that processes sequences in forward and backward direction to consider the influence of timesteps in both directions is employed. For further learning, unidirectional layers’ stacking is implemented. The performance of the model is optimized by fine-tuning hyperparameters, regularization techniques for overfitting resolution, and various merging options for the bidirectional input layer. The proposed model achieves good optimization and performs better than the simple LSTM and a Recurrent Neural Network (RNN) based model. Moreover, an attention-based mechanism is adopted to focus on more significant timesteps for prediction. The self-attention approach improves performance further and works well especially for longer sequences and extended time horizons. Experiments are conducted using real-world data collected, and results are evaluated using the mean square error loss function.

Parole chiave

  • Air quality forecasting
  • Air pollution forecasting
  • Deep learning
  • Long short term memory
  • attention
Accesso libero

ESAR, An Expert Shoplifting Activity Recognition System

Pubblicato online: 10 Apr 2022
Pagine: 190 - 200

Astratto

Abstract

Shoplifting is a troubling and pervasive aspect of consumers, causing great losses to retailers. It is the theft of goods from the stores/shops, usually by hiding the store item either in the pocket or in carrier bag and leaving without any payment. Revenue loss is the most direct financial effect of shoplifting. Therefore, this article introduces an Expert Shoplifting Activity Recognition (ESAR) system to reduce shoplifting incidents in stores/shops. The system being proposed seamlessly examines each frame in video footage and alerts security personnel when shoplifting occurs. It uses dual-stream convolutional neural network to extract appearance and salient motion features in the video sequences. Here, optical flow and gradient components are used to extract salient motion features related to shoplifting movement in the video sequence. Long Short Term Memory (LSTM) based deep learner is modeled to learn the extracted features in the time domain for distinguishing person actions (i.e., normal and shoplifting). Analyzing the model behavior for diverse modeling environments is an added contribution of this paper. A synthesized shoplifting dataset is used here for experimentations. The experimental outcomes show that the proposed approach attains better consequences up to 90.26% detection accuracy compared to the other prevalent approaches.

Parole chiave

  • Automated surveillance system
  • Human Activity Recognition (HAR)
  • Histogram of Oriented Gradient (HOG)
  • Optical Flow
  • Convolutional Neural Network (CNN)
  • Long Short Term Memory (LSTM)

Pianifica la tua conferenza remota con Sciendo