International Journal of Business Intelligence and Data Mining (24 papers in press)

Regular Issues

• Brain Hemorrhage Classification from CT Scan Images using Fine-tuned Transfer Learning Deep Features  
  by Arpita Ghosh, Badal Soni, Ujwala Baruah 
  Abstract: Classification of brain haemorrhage is a challenging task and needs to solved to help advance medical treatment. Recently, it has been observed that efficient deep learning architectures have been developed to detect such bleeding accurately. The proposed system includes two different transfer learning strategies to train and fine tune ImageNet pre-trained state-of-the-art architecture such that VGG 16, Inception V3, DenseNet121. The evaluation metrics have been calculated based on the performance analysis of the employed networks. Experimental results show that the modified fine-tuned Inception V3 perform well and achieved the highest test accuracy.
  Keywords: transfer learning; VGG 16; Inception V3; DenseNet121; brain haemorrhage; ReLU; binary cross entropy.
  DOI: 10.1504/IJBIDM.2022.10046012
   
  
• Fuzzy Twin Kernel Ridge Regression Classifiers for Liver Disorder Detection  
  by Deepak Gupta, Barenya Bikash Hazarika, Parashjyoti Borah 
  Abstract: The liver is a key organ in the human body that aids in the digestion of food, the elimination of toxins, and the storage of energy. Patients with liver disorders are on the rise all over the world. However, because the disorders symptoms are unclear, it is difficult to diagnose it, which raises the diseases death rate. The study introduces novel fuzzy twin models for liver disease classification. In the first model, the membership is calculated based on the quadratic function called fuzzy twin kernel ridge regression-quadratic (FTKRR-Q). In the second model, we have calculated the fuzzy membership based on the centroid and named the model as fuzzy twin kernel ridge regression-centroid (FTKRR-C). For our research, the BUPA or liver disease dataset has been used from the UCI machine learning repository. Experimental results are compared with the twin support vector machine, kernel ridge regression classifier and twin kernel ridge regression classifier. The accuracy, sensitivity, F1-score, and Mathews correlation coefficient are used to evaluate the suggested models performance. Experiments are also carried out on some real-world benchmark datasets. The results reveal the applicability of the proposed models.
  Keywords: Twin kernel ridge regression; Fuzzy membership; Liver disorder; Biomedical data; Classification.
  DOI: 10.1504/IJBIDM.2024.10052716
   
  
• Temporal Autoencoder Architectures with Attention for ECG Anomaly Detection  
  by Ann Varghese, Midhun MS, James Kurian 
  Abstract: Anomaly detection is a crucial step in any diagnostic procedure. With the advent of continuous monitoring devices, it is inevitable to use technological assistance for the same. Many methods, including autoencoders, have been proposed for anomaly detection in time series ECG data. The attention mechanism dynamically highlights the relevant portion of the input data and provides the decoder with the information from every encoder hidden state in its temporal vicinity. This work proposes a performance enhancement of autoencoders in identifying an ECG anomaly with the help of attention. A comparison of different autoencoder models, LSTM and hybrid, with and without attention to detect an anomaly, is proposed in this work. The comparison of the different models in terms of precision, recall, F1-score, false-positive rate (FPR), false-negative rate (FNR) and area under the ROC curve (AUC) are specified. The obtained results indicate that attention helps to enhance the autoencoder’s performance.
  Keywords: arrhythmia; hybrid; long-short-term memory; LSTM; convolution; MIT-BIH; time-series.
  DOI: 10.1504/IJBIDM.2024.10052839
   
  
• Challenges and Issues in Facial Emotion Recognition Techniques  
  by Nizamuddin Khan, Dr. Ajay Vikram Singh, Rajeev Agrawal 
  Abstract: We analyse the most obvious distributed writing in Facial emotion recognition conveyed throughout the last decade in this paper. This paper assesses the analysis of the work done as such far and tries to assess each based on a set of composite parameters. Also, under the equivalent, bearing is looking for future work based on new specialisation regions and identifying research gaps. There is a project underway to determine the optimum procedure that could be employed to meet all of the established constraints. Furthermore, a heading is expected for work ahead on new stronghold locations below the same. The magnitude of such labour cannot be undervalued, especially for the injured and the more prepared, where outer appearance may continue to play an important role in conveying thoughts in a robotic manner. Furthermore, there are numerous other beneficial applications that cover a wide variety of our daily lives.
  Keywords: facial action coding system; FACS; convolutional neural network; CNN; feature expression recognition; FER.
  DOI: 10.1504/IJBIDM.2024.10053050
   
  
• Innovative Model of Real Estate Financing Based on Internet of Things Thinking  
  by Shihan Zhang 
  Abstract: This paper studied the model innovation of real estate financing and proposed the financing model of the internet of things platform. From the five aspects of financing cost, financing efficiency, financing information transparency, financing risk-taking level and enterprise evaluation, the experimental study was carried out on the internet of things platform financing model and the traditional real estate financing model. The research showed that the financing model of the internet of things platform could improve the financing efficiency and reduce the level of financing risk.
  Keywords: IoT thinking; real estate financing; financing model innovation; traditional real estate financing model; IoT platform financing model.
  DOI: 10.1504/IJBIDM.2024.10054654
   
  
• Data Augmentation and Denoising of Computed Tomography (CT) Scan Images In Training Deep Learning Models For Rapid COVID-19 Detection  
  by Auwalu Mubarak, Sertan Serte, Fadi Al-Turjman, Zubaida Said Ameen 
  Abstract: The deadly respiratory disease Corona Virus-2 (COVID-19) which was declared a pandemic by the World Health Organization (WHO) has resulted in over a million death around the world within less than a year With the rapid spread of the virus, the currently adopted COVID-19 test by the WHO is the reverse transcription polymerase chain reaction (RT-PCR) test, which is expensive, time-consuming and not accessed by underdeveloped countries computed tomography (CT) scan images that were used in profiling suspected COVID-19 patients can serve as an alternative to the RT PCR test method In this study, two different pre-trained deep learning models ResNet-50 and ResNet-101 were trained to classify positive COVID-19 scan images The best model which was trained on the augmented CT scan images achieved an accuracy of 98.3%, a sensitivity of 0.984, specificity of 0.983.
  Keywords: artificial intelligence; deep learning; CT scan images; medical imagine; transfer learning; data augmentation; denoising.
  DOI: 10.1504/IJBIDM.2024.10055417
   
  
• The application of Machine Learning in Real Estate Enterprise Risk Management  
  by Hongtao Pan 
  Abstract: As a pillar industry of the national economy, real estate plays a pivotal role in economic development, social stability, employment and other fields. In recent years, the rapid development of the real estate industry has become a common concern of the whole society. The changes of the market economy and the national macro-control will have a huge impact on the real estate industry, and the real estate industry has become the industry with the highest risk. It is an inevitable trend to introduce new technologies into the risk management of real estate enterprises and to improve the automation and intelligence of risk management. By analysing the risk management of real estate enterprises, this paper constructs a risk management model and implementation path based on machine learning, and elaborates its implementation process in detail. The study found that compared with the risk level of the real estate industry, under the enterprise risk management model based on machine learning, the risk level of enterprises is lower.
  Keywords: risk management; real estate business; machine learning; business management.
  DOI: 10.1504/IJBIDM.2024.10055528
   
  
• Biomedical Signal to Image Conversion and Classification using Flexible Deep Learning Techniques  
  by Abhishek Das, Soumya Ranjan Nayak, Mihir Mohanty, Piyush Kumar Shukla 
  Abstract: Most of the diseases are diagnosed from the image data like CT scans, MRIs, and X-rays. Gene data carries vital information related to the diseases that needs to be analysed for diagnosis. Both the logics are combined and a flexible deep ensemble learning-based model is proposed for the classification of images generated from one-dimensional (1D) data. Earlier works in the detection of brain tumours and epileptic seizures have been developed either directly providing 1D data or images to the classification model, whereas the proposed method utilises the effectiveness of two-dimensional (2D) convolutional neural networks (CNNS) to analyse 1D data like gene expressions and EEG signals after effective conversion to images. The data conversion is performed using three data reduction techniques, i.e., locally linearly embedding (LL-Embedding), multi-dimensional scaling (MDS), and t-distributed stochastic neighbour embedding (t-SNE) with convex hull algorithm to wrap all the data points. Multilayer perceptron is used for second-stage training. The proposed method is verified using brain tumour gene data collected from the genomic data commons (GDC) data portal and the EEG data for epileptic seizures detection provided by the University of Bonn (UoB dataset) and provided 97.38% and 97.33% accuracies respectively.
  Keywords: gene to image; EEG to image; convolution neural network; ensemble learning; multilayer perception; brain tumour; epileptic; seizure; convex hull.
  DOI: 10.1504/IJBIDM.2024.10055529
   
  
• Intelligent Construction Project Management Schedule Technology Based on Internet of Things Computing  
  by Xiao Wang 
  Abstract: In recent years, the development of the information technology industry has driven changes in all walks of life. Under the background of information technology, the construction industry has also achieved transformation and development, which is mainly reflected in the increase of modern building intelligent system engineering. Intelligent system management is different from ordinary construction projects, its own complexity and systematicness exceed ordinary projects. Project management schedule technology is an effective management model for intelligent building system engineering projects. Project management schedule can effectively control project cost and project completion quality, effectively ensure the delivery of construction schedule, and effectively reduce construction cost and resource consumption. This article looks at construction project management, pointing out that IoT assistance to current project progress can improve the overall quality of smart buildings. The experiment shows that the internet of things has increased the efficiency of construction project management by 72.73%.
  Keywords: internet of things computing; intelligent building; intelligent building engineering; construction project; project management progress.
  DOI: 10.1504/IJBIDM.2024.10055705
   
  
• An intelligent Recommendation Method of Remote Ideological and Political Education Resources Based on User Clustering  
  by Yuan Zhang 
  Abstract: Aiming at the remote ideological and political education resources, due to the problems of low recall, poor recommendation effect and long recommendation time in traditional methods, a new intelligent recommendation method is proposed. First, user behaviour data with HTML5 and JavaScript technology were collected, and then filtering, dimension reduction, redundancy elimination and missing completion operations on the collected data were implemented. Then, users were modelled through the FBP model and user similarity was calculated. Finally, users with similar preferences were clustered using k-means++algorithm, the class of target users was searched, a scoring prediction matrix was established, and the highest scoring remote ideological and political education resources to users were recommended. The experimental results show that the proposed method cannot improve the recommendation accuracy and recall, but also shorten the recommendation time.
  Keywords: user clustering; distance ideological; political education; user similarity; k-means++algorithm; scoring prediction matrix.
  DOI: 10.1504/IJBIDM.2024.10055797
   
  
• Detection method of students' online learning state based on posture recognition  
  by Xiaowei He 
  Abstract: Because of the problems of low detection accuracy and long detection time in traditional online learning state detection methods, a new method based on posture recognition is proposed. First of all, a pinhole camera perspective imaging model is constructed, students’ online learning images are collected, and the images are processed with greyscale, smoothing, enhancement and light compensation. Secondly, according to the key points of bones, the online learning image features of students after preprocessing are extracted. Finally, identify students’ online learning posture, and construct a state detection model combining eye movement behaviour to complete the detection of students’ online learning state. The experimental results show that the proposed method has higher accuracy and shorter detection time for students’ online learning state detection.
  Keywords: posture recognition; online learning status; Gaussian filtering; Laplace operation; bone keys.
  DOI: 10.1504/IJBIDM.2024.10055839
   
  
• A Multi agent interactive teaching effect evaluation method based on matrix method  
  by Tiantian Xie 
  Abstract: In order to overcome the problems of low assessment value and high assessment error in traditional methods, this paper proposes a multi-agent interactive initiating effect assessment method based on matrix method. Firstly, each initiating subject to construct an interactive assessment carpal was considered. Secondly, the correlation degree matrix is constructed by matrix method, the initiating effect assessment matrix is determined, and its product sum formula is solved. The least square method of weight is used to calculate the index weight, construct the initiating effect assessment function, determine the initiating effect assessment grade, output the initiating effect assessment results, and realise the multi-subject interactive initiating effect assessment. The results show that the calculation result of effect assessment is 2.124, the scoring value is 86, and the assessment error is not more than three points, which shows that this method can improve the assessment effect.
  Keywords: matrix method; interactive; initiating effect assessment; correlation degree matrix; the least square method of weights; index weight.
  DOI: 10.1504/IJBIDM.2024.10055840
   
  
• An automatic error correction method for business English text translation based on natural language processing  
  by Yan Yang 
  Abstract: In order to improve the efficiency and accuracy of traditional translation error correction methods, this paper proposes an automatic error correction method for business English text translation based on natural language processing. First, the error correction statements are divided into triples by using natural language transformation to determine the logical expression template. Secondly, maximum likelihood is used to estimate word frequency. Then, a naive Bayesian classifier is constructed to classify translation words. Finally, beam search decoding is used to generate target text correction sentences, and an automatic error correction function for text translation is constructed. The results show that the accuracy of translation error correction can reach 99%, the average error of error correction is less than 0.10, and the error correction time is only 4.5 s. This method can improve the accuracy and efficiency of error correction.
  Keywords: natural language processing; loss function; business English translation; Markov assumption; naive Bayes.
  DOI: 10.1504/IJBIDM.2024.10056121
   
  
• Research on Classification of Educational Digital Resources Based on KNN Algorithm  
  by Zizhen Xiao 
  Abstract: In order to overcome the problems of low classification accuracy and long classification time in traditional educational digital resources classification methods, a classification method of educational digital resources based on KNN algorithm is proposed. First, the crawler technology is used to automatically obtain educational digital resource data on the web page through the search engine, and store it in the resource library in a distributed manner. Then, the obtained educational digital resource data is pre-processed through cleaning, word segmentation and stop word removal. Then, according to the processing results, combined with chi square statistics and PCA methods, the characteristics of educational digital resources are selected twice. Finally, the results are selected according to the characteristics, K-nearest neighbour classification (KNN) algorithm is used to classify educational digital resources. The simulation results show that the proposed method has higher accuracy and shorter classification time.
  Keywords: KNN algorithm; educational digital resources; crawler technology; Chi square statistical method; PCA method.
  DOI: 10.1504/IJBIDM.2024.10055841
   
  
• Research on anomaly recognition of the English MOOC teaching platform based on deep feature learning  
  by Haixia YU 
  Abstract: To improve the recognition rate and accuracy of the English MOOC teaching platform access exceptions, and reduce the recognition time, the anomaly recognition method of the English MOOC teaching platform based on deep feature learning is proposed. Firstly, analyse the abnormal data accessed by English MOOC teaching platform under the influence of Gaussian white noise, and denoise the abnormal data. Then, according to the data denoising results, the features of abnormal data accessed by English MOOC teaching platform are extracted. Finally, using depth feature learning, unsupervised pre training stack noise reduction self-encoder layer by layer obtains the depth features of abnormal data, and combined with softmax classifier to realise the recognition of abnormal access to English MOOC teaching platform. Experimental results show that the anomaly recognition rate and accuracy of the proposed method are 98.1% and 96.8% respectively, and the anomaly recognition time is only 4.7 s.
  Keywords: deep feature learning; stack noise reduction self-encoder; English MOOC; softmax classifier; teaching platform; access exception identification.
  DOI: 10.1504/IJBIDM.2024.10055842
   
  
• Push Method of Online Learning Resources Based on User Behavior Characteristics  
  by Liu Xiangyuan 
  Abstract: Aiming at the low accuracy of user behaviour feature extraction and resource recommendation in online learning resource push, an online learning resource push method based on user behaviour feature is proposed. First, XGBoost model is constructed, and user feature data is extracted by combining decision tree. Then, a graph convolution neural network model is constructed to preprocess user characteristic data. Finally, K-means algorithm is introduced to build online learning resource recommendation model based on user feature data to achieve resource recommendation. The experimental results show that the user feature extraction accuracy of the proposed method is higher than 95%, the recommendation accuracy is 96%, and the recommendation time cost is less than 0.5s, which improves the recommendation effect.
  Keywords: user behaviour characteristics; online learning resources; push; XGBoost model; customer characteristic similar data.
  DOI: 10.1504/IJBIDM.2024.10055843
   
  
• Prediction Method of MOOC Teaching Effect Based on Data Mining  
  by Aiping Wang 
  Abstract: Aiming at the problems of low prediction accuracy and large prediction time of MOOC teaching effect, a method of MOOC teaching effect prediction based on data mining is proposed. Firstly, this paper analyses the learning behaviour of the course, obtains the learning characteristics, determines the type of learning behaviour data, completes the learning behaviour analysis based on the clustering algorithm, and finds all the learning behaviour data within the limited range. Then, by analysing the characteristics of MOOC teaching mode, we determine the teaching effect prediction index and build the effect prediction index system. Finally, based on the integrated learning algorithm, a multi-classifier is constructed to calculate the weight of the prediction index, and the MOOC teaching effect prediction model is constructed to complete the final prediction of the teaching effect. The test results show that the prediction accuracy of the proposed method is higher than 95%, and the maximum prediction time cost is 4 s, which can effectively improve the prediction accuracy, shorten the time cost, and have a good prediction effect.
  Keywords: data mining; MOOC teaching effect; course characteristics; integrated learning; constraint function.
  DOI: 10.1504/IJBIDM.2024.10055844
   
  
• Personalized recommendation of educational resources based on collaborative filtering  
  by Yi Song 
  Abstract: In order to overcome the problems of low accuracy and long time of personalised recommendation of educational resources in traditional personalised recommendation methods of educational resources, a personalised recommendation method of educational resources based on collaborative filtering is proposed. First, calculate the membership relationship between knowledge points and educational resources, then model educational resources with the help of fuzzy set theory, model learners through four models, then establish learner interest matrix, extract learner characteristics, finally, calculate learner similarity based on learner collaborative filtering recommendation algorithm, and select the latest learner set. The score of the nearest neighbour learner on the item is used to predict the score of the target learner, so as to generate recommendation results. The simulation results show that the proposed method has higher accuracy and shorter recommendation time for personalised recommendation of educational resources.
  Keywords: collaborative filtering; educational resources; interest matrix; personalised recommendation; similarity.
  DOI: 10.1504/IJBIDM.2024.10055845
   
  
• Personalized recommendation method of online ideological and political education resources in colleges and universities based on spectral clustering  
  by Xixi Zhang  
  Abstract: Due to the problems of low accuracy and low F value in online educational resource recommendation using existing methods, a personalised recommendation method for online educational resource recommendations based on spectral clustering is proposed for ideological and political teaching in universities. First, Pearson formula was used to determine the degree of similarity and collect data from hidden online ideological and political education resources. Then, Kalman filtering algorithm is used to preprocess the collected relevant resource data. Finally, the spectral clustering algorithm is used to determine the data distance of resources according to the neighbour relationship, so as to make the recommended resources personalised, so as to build the personalised recommendation model for online educational resources, and realise the recommendation of resources. The experimental results show that the F value of the proposed method is about 1.0 and the paste progress and recommendation accuracy are both higher than 95%, which are 10% and 12% higher than the comparison method respectively, which verifies the good recommendation effect of this method.
  Keywords: spectral clustering; ideological and political education; online education resources; personalised recommendation; objective function; minimum cut set.
  DOI: 10.1504/IJBIDM.2024.10056141
   
  
• A personalized push method of English mobile reading resources based on tag similarity  
  by Juanzhi Shi, Yadan Deng 
  Abstract: Aiming at shortening the recommendation time of reading resources and improving the retention rate of recommendation results, a new personalised push method of English mobile reading resources was studied based on label similarity. Firstly, preprocess resource data by data cleaning and noise reduction. K-means clustering algorithm is used to cluster resource data. Secondly, calculate the frequency of resource tags used by users according to their implicit needs and preferences. Finally, the cosine similarity between user tags and resource tags in the same category is obtained, and a push list is formed to push resources before ranking test for users. The experimental results show that the maximum generation time of the push list is only 4,372 ms after the application of this method, and the retention rate of the recommendation results is always above 93%.
  Keywords: tag similarity; user tag; resource tag; English mobile reading resources; personalised push; K-means clustering; push list.
  DOI: 10.1504/IJBIDM.2024.10056142
   
  
• Knowledge mapping based online teaching resource recommendation method for chinese education  
  by Huishuang Qi 
  Abstract: In order to overcome the problems of poor recommendation accuracy, low recommendation recall and long recommendation time in traditional teaching resource recommendation methods, a knowledge map based Chinese online teaching resource recommendation method is proposed. First, the TF-IDF method is used to calculate the weighted value of Chinese education recommended resources, and then the ontology of the knowledge map of Chinese education online teaching resources is designed to obtain the representation of the characteristics of teaching resources. Finally, the cosine similarity method is used to calculate the similarity and interest of user attribute vectors, and the predicted interest of users in resources is obtained. According to the predicted interest value, the recommended list of cultural education online teaching resources is generated from high to low, Get recommended results. The experimental results show that this method has a better effect on online teaching resources recommendation of Chinese education.
  Keywords: cosine similarity; knowledge graph; TF-IDF method; resource recommendation; Chinese education.
  DOI: 10.1504/IJBIDM.2024.10056143
   
  
• Design and Application of a New Massive Open Online Training Platform  
  by Zhe Wang, Jingjing Li 
  Abstract: Over the past decade, e-learning has been a tremendous success. Many MOOCs have been established around the world. A mounting number of college students study through MOOCs. But with the rapid development of cloud computing, big data and artificial intelligence technology, there are an increasing number of IT training requirements in colleges. The mode of traditional IT training can not match the needs of college students and teachers, such as fragmentation, anytime, anywhere, etc. IT training is more complex than e-learning, so it cannot simply copy the mode of e-learning. This article proposes a framework of a massive open online training platform based on docker and cloud services, which overcomes the limitations of traditional training. It has the advantages of quickness, economy, efficiency and flexibility. It provides one-stop services for online learning, online training and online assessment for college students. The application results of Shuishan online show that the proposed framework is reliable and excellent.
  Keywords: e-learning; massive open online training platform; MOOTP; online training; cloud services; Docker.
  DOI: 10.1504/IJBIDM.2024.10058390
   
  
• Visual Analysis of Financial Big Data Based on Data Center  
  by Wentao Yang, Rui Tian, Lifang Zhang 
  Abstract: This paper aims to study the data middle-end means in data mining (DM) technology, and combine back propagation (BP) network to research and analyse the visualisation of financial data, so that financial data can be better combined with intelligent and convenient means. Based on the experiments in this paper, it can be seen that by analysing the relevant financial statements of H Group from 2016 to 2019, it can be effectively found that the accuracy of financial data has been effectively improved in 2016 after combining the data centre. The experimental results of this paper have shown that using the BP network as the basic method to study the visual analysis of financial big data can obtain intuitive and scientific relevant experimental data, and help the managers of the management organisation to make scientific decisions.
  Keywords: data middle office; financial data; finance visualisation; data mining; BP neural network.
  DOI: 10.1504/IJBIDM.2024.10060196
   
  
• Information Accuracy Verification System of Invoice Metadata Based on Data Warehouse  
  by Hui Chen, Yuening Wang, Weilin Xue 
  Abstract: This paper aims to explore the information accuracy verification system based on the invoice metadata of the data warehouse. Metadata defines the role of the data warehouse and specifies the content and location of the data warehouse, and describes the data extraction and transformation rules. The application of neural network in data warehouse was discussed. Finally, based on this research, it analysed the management and application of electronic invoice metadata. The experimental results of this paper showed that to accelerate the popularisation of electronic invoices and vigorously promote electronic invoices is an inevitable choice for developing the economy and deepening tax reform. When processing electronic invoice filing, it is necessary to realise the automatic entry of core metadata in the electronic document management system, so that it has the function of core metadata and gives full play to its due role.
  Keywords: Electronic Invoice; Data Warehouse; BP Neural Network; Invoice Metadata.
  DOI: 10.1504/IJBIDM.2024.10060201