Development Of Data Mining Technology Research Articles

A Dynamic Multi-Scale Convolution Model for Face Recognition Using Event-Related Potentials.

With the development of data mining technology, the analysis of event-related potential (ERP) data has evolved from statistical analysis of time-domain features to data-driven techniques based on supervised and unsupervised learning. However, there are still many challenges in understanding the relationship between ERP components and the representation of familiar and unfamiliar faces. To address this, this paper proposes a model based on Dynamic Multi-Scale Convolution for group recognition of familiar and unfamiliar faces. This approach uses generated weight masks for cross-subject familiar/unfamiliar face recognition using a multi-scale model. The model employs a variable-length filter generator to dynamically determine the optimal filter length for time-series samples, thereby capturing features at different time scales. Comparative experiments are conducted to evaluate the model's performance against SOTA models. The results demonstrate that our model achieves impressive outcomes, with a balanced accuracy rate of 93.20% and an F1 score of 88.54%, outperforming the methods used for comparison. The ERP data extracted from different time regions in the model can also provide data-driven technical support for research based on the representation of different ERP components.

Research on learning behavior patterns from the perspective of educational data mining: Evaluation, prediction and visualization

The rapid growth of educational data creates the requirement to mine useful information from learning behavior patterns. The development of data mining technology makes educational data mining possible. The paper intends to use a public educational data set to study learning behavior patterns from the perspective of educational data mining, so as to promote the innovation of educational management. Firstly, in order to reduce the dimension of data analysis that facilitates the improvement in efficiency, principal component analysis is carried out to reduce the number of attributes in the data set. The significant attributes in the rotating principal component matrix rather than principal components which are not closely related to learning behavior patterns are extracted as the research variables. Then, a pseudo statistic is proposed to determine the number of clusters and the preprocessed data set is clustered according to the extracted attributes. The clustering results are applied to add class labels to the data, which is convenient for the later data training. Finally, six classification algorithms J48, K-Nearest Neighbor, Bayes Net, Random Forest, Support Vector Machine and Logit Boost are used to train the data with labels and build prediction models. At the same time, the performance and applicable conditions of six classifiers in terms of accuracy, efficiency, error, and so on are discussed and compared. It is found that the performance of the integrated algorithm is better than that of a single classifier. In the integrated algorithm, compared with Random Forest, the running time of Logit Boost is shorter.

A framework of genealogy knowledge reasoning and visualization based on a knowledge graph

PurposeWith the development of data mining technology, diverse and broader domain knowledge can be extracted automatically. However, the research on applying knowledge mapping and data visualization techniques to genealogical data is limited. This paper aims to fill this research gap by providing a systematic framework and process guidance for practitioners seeking to uncover hidden knowledge from genealogy.Design/methodology/approachBased on a literature review of genealogy's current knowledge reasoning research, the authors constructed an integrated framework for knowledge inference and visualization application using a knowledge graph. Additionally, the authors applied this framework in a case study using “Manchu Clan Genealogy” as the data source.FindingsThe case study shows that the proposed framework can effectively decompose and reconstruct genealogy. It demonstrates the reasoning, discovery, and web visualization application process of implicit information in genealogy. It enhances the effective utilization of Manchu genealogy resources by highlighting the intricate relationships among people, places, and time entities.Originality/valueThis study proposed a framework for genealogy knowledge reasoning and visual analysis utilizing a knowledge graph, including five dimensions: the target layer, the resource layer, the data layer, the inference layer, and the application layer. It helps to gather the scattered genealogy information and establish a data network with semantic correlations while establishing reasoning rules to enable inference discovery and visualization of hidden relationships.

Performance Evaluation of Machine Learning Methods for Heart Failure Prediction

Heart failure is a syndrome of cardiac circulation disorder. Due to the dysfunction of the systolic function or diastolic function of the heart, the venous blood volume cannot be fully discharged from the heart, resulting in blood stasis in the venous system and insufficient perfusion in the arterial system. The symptoms of this disorder are concentrated in pulmonary congestion and vena cava congestion. The correlation between the inducement of heart failure and the incidence of heart failure is a subject that needs to be studied in the medical field. In recent years, with the development of data mining technology, more and more analytical models and algorithms have been applied in the medical field, which greatly improve the efficiency of medical data analysis and enable medical workers to cure diseases better. In this study, an ensemble learning model is applied to analyze the data of heart failure. First, the data is preprocessed and normalized, and features that are not associated with death rate of heart failure are removed. Secondly, multiple base classifiers are trained and compared. Finally, the competent base classifiers are selected and integrated with the Stacking-based ensemble learning algorithm for final classification. Comparative analysis showed that the prediction results of ensemble model are better than that of base classifiers in evaluation indexes such as accuracy, precision, AUC, Balanced accuracy and F1-score for the heart failure data.

Marketing Decision Support System Based on Data Mining Technology

With the continuous development of business intelligence technology, the application research of decision support systems (DSSs) is deepening. In China, the work in this area started relatively late, and there are few DSS research cases to assist in marketing decision-making. Currently, marketing decision support systems have shortcomings in data integration, historical data, query functions, and data analysis. This article analyzes the characteristics of marketing decision-making, discusses the application of data warehouse, OLAP, and data mining technology in marketing decision support systems, and designs a marketing decision support system based on data mining technology. The system uses a BP neural network to conduct data mining marketing forecasting. A three-layer network model for marketing prediction is established, with sales time, product price, and customer purchasing power as network inputs and output as the sales volume of a certain type of product in different locations. The test results show that the average absolute percentage error of this method is 15.13%, and the prediction accuracy is high. Research shows that with the continuous development of data mining technology, the system cannot only help users conduct scientific and reasonable marketing decision-making analyses, making the marketing decision-making process more scientific and reasonable, but also can bring new ideas to enterprise decision-makers, promoting the continuous improvement and progress of the system.

Optimized Apriori algorithm for deformation response analysis of landslide hazards

With the explosive development of data mining technology, higher requirements have been put forward to analyze the response of landslide deformation. However, related algorithms, such as the traditional Apriori and FP-Growth algorithms, are still in the starting period of being applied to landslide hazards. Due to monitoring data characteristics, some problems were encountered while applying these algorithms, such as poor applicability and low computational efficiency. Therefore, we propose an optimized Apriori algorithm to solve the above problems. The optimized algorithm strictly controls the front and rear itemsets' construction process and stores the factors and deformation events according to their dimensions and levels. In addition, some key calculation processes of the algorithm are well-parallelized. Based on the monitoring data of the Baishuihe landslide, three experiments were designed to verify the performance of the proposed algorithm. The results show that when the strong association rules with high factor dimension and level characteristics were obtained, the proposed algorithm's computation time was 1/432 and 1/80 of the Apriori and FP-Growth algorithms, respectively. The proposed algorithm has significant advantages in analyzing massive high-dimensional monitoring data of landslide hazards.

Analysis of Influencing Factors and Enterprise Strategy of Online Consumer Behavior Decision Based on Association Rules and Mobile Computing

Consumers are constantly generating a large amount of data, thanks to the arrival of the big data era and the advancement of mobile edge computing capabilities. Massive behavioral data points to the need to mine and analyze potentially valuable information. The commonality and individuality of customer groups’ consumption behaviors must be researched before marketing decisions and strategies can be implemented. Because of the unique advantages of mobile edge computing technology, the Internet’s application has become more and more widespread, and businesses are increasingly paying attention to network marketing. With the system’s long-term use, decision-makers began to wonder if useful information could be extracted from vast amounts of historical data to help them summarize or even predict changes in customer demand and purchasing behavior. This assumption is possible thanks to the rise and development of data mining technology. Association rules are increasingly being applied to customer behavior analysis as the most active branch of data mining in the last ten years. The majority of association rule research currently focuses on one-dimensional data association analysis of a user’s package using classical algorithms. The use of association rules mining on multidimensional data with multiple attributes in the telecom service industry is limited due to the complexity of the data structure and algorithm.

Design of enterprise human resources decision support system based on data mining

The management and development of human resources have become one of the core contents of enterprise management. It is from the initial use as an auxiliary tool in human resources business management to the later development of a relatively independent and functional sub-system for the processing of personnel matters between modules, and then to achieve the integration of various sub-systems to form a relatively complete set of human resources information management system. In recent years, with the rapid development of data mining technology, more and more enterprises have begun to attach importance to the assistance of computer information systems to manage the completion of various decision-making activities in operation. This paper designs an enterprise human resource decision support system based on data mining to more effectively support human resource management business information, data standardization, intelligent decision analysis, optimize human resource business processes, improve human resource management business efficiency, provide more effective decision support for enterprise operation and promote enterprise development. Results of the study show the effectiveness of the research.

Fuzzy Data Mining Technology Based on Time Sequence in Archives Management System

The rapid development of data mining technology enables the processing of unstructured data in archives. It can help realize the multiple classification of archives, the accurate retrieval of archive information, the integrated presentation of archive content, and the scientific standardization of archive appraisal. This article mainly introduces the research of fuzzy data mining technology based on time series in archives management system. This paper uses the research on the time series-based fuzzy data mining technology in the archives management system, uses the data mining technology to plan and design the archives management system, and combines the relevant procedures of the archives management work to discuss what should be followed when designing the archives data mining model. It also introduces the basic framework of archive data mining in detail, mainly from the aspects of database construction, data preprocessing, data document establishment, training process design, application process architecture and so on. The experimental results of this paper show that the research on the time series-based fuzzy data mining technology in the file management system has increased the efficiency of file management by 19%. The limitation of the research on the time series-based fuzzy data mining technology in the file management system is the data Mining applications provide good methods and ways to analyze, discuss and summarize, so as to enrich the academic research results.

Predicting Student Achievement Based on Temporal Learning Behavior in MOOCs

With the development of data mining technology, educational data mining (EDM) has gained increasing amounts of attention. Research on massive open online courses (MOOCs) is an important area of EDM. Previous studies found that assignment-related behaviors in MOOCs (such as the completed number of assignments) can affect student achievement. However, these methods cannot fully reflect students’ learning processes and affect the accuracy of prediction. In the present paper, we consider the temporal learning behaviors of students to propose a student achievement prediction method for MOOCs. First, a multi-layer long short-term memory (LSTM) neural network is employed to reflect students’ learning processes. Second, a discriminative sequential pattern (DSP) mining-based pattern adapter is proposed to obtain the behavior patterns of students and enhance the significance of critical information. Third, a framework is constructed with an attention mechanism that includes data pre-processing, pattern adaptation, and the LSTM neural network to predict student achievement. In the experiments, we collect data from a C programming course from the year 2012 and extract assignment-related features. The experimental results reveal that this method achieves an accuracy rate of 91% and a recall of 94%.

Multi-view SVM Classification with Feature Selection

With the rapid development of data mining technology, multi-view learning (MVL) has become a new research field, which has attracted wide attention of scholars at home and abroad. Multi-view learning is to combine multiple view data of the same entity for data classification, thereby improving learning performance. Previous multi-view research methods mainly concentrate on the relationship between different data views for classification problems. However, when the data is in high dimensions, it is necessary to perform feature selection in the multi-view data classification process. In this paper, we proposed a Multi-view Support Vector Machine Classification with Feature Selection (MSVMCFS) algorithm, which can not only classify multi-view data, but also select features for each view data in the process of classification. In the model, feature selection is performed by the l1 norm sparsity regularization, and consistency and complementarity between the two views are maintained. To achieve the optimization goal, we adopt linear programming to solve the model. The experimental results on 30 binary datasets demonstrate the validity of the model.

A Method for Repairing Process Models Containing a Choice With Concurrency Structure by Using Logic Petri Nets

Process models are used to describe or verify the correctness of business processes before their implementation by enterprise information systems. With the development of data mining technologies, a lot of event logs generated from such systems can be used to check conformance of business processes and process models. For models containing a choice with concurrency structure, although the model repaired by the existing method can replay the event log, it is quite different from the business process. Therefore, we propose a model repair method based on logic Petri net. First, we give an algorithm to find the position of the deviation. Then, we add directed arcs from a place to a transition or from a transition to a place via logic Petri nets. Finally, we repair the model by adding logic transitions with logic functions. The repaired model can efficiently enhance conformance dimensions, such as simplicity and precision. The correctness and effectiveness of the method are verified by experimental results.

Association Rule Mining for Selecting Proper Students to Take Part in Proper Discipline Competition: A Case Study of Zhejiang University of Finance and Economics

In recent years, the educational issues have attracted more and more researchers’ and teachers’ attention. On the other hand, the development of data mining technology, provides a new method to extract the useful information from the complex educational data. In order to increase the chance of students to be awarded in discipline competition, it is better to select the proper students to take part in the proper discipline competition. Therefore, in this study, we collect the information of 164 undergraduate students as a case study. All students majored in Software Engineering in Zhejiang University of Finance and Economics. The Apriori algorithm with group strategy is used to find the relationship between the students’ courses scores and competition awards. According to the results of association rule mining, we find that the students with higher scores of C# Development, Object-Oriented, Internet Web Design, Data Structure(C#), and Basic Programming will have a higher probability to be awarded in the competition.

Research on PM2.5 time series characteristics based on data mining technology

With the development of data mining technology and the establishment of environmental air quality database, it is necessary to discover the potential correlations and rules by digging the massive environmental air quality information and analyzing the air pollution process. In this paper, we have presented a sequential pattern mining method based on the air quality data and pattern association technology to analyze the PM2.5 time series characteristics. Utilizing the real-time monitoring data of urban air quality in China, the time series rule and variation properties of PM2.5 under different pollution levels are extracted and analyzed. The analysis results show that the time sequence features of the PM2.5 concentration is directly affected by the alteration of the pollution degree. The longest time that PM2.5 remained stable is about 24 hours. As the pollution degree gets severer, the instability time and step ascending time gradually changes from 12-24 hours to 3 hours. The presented method is helpful for the controlling and forecasting of the air quality while saving the measuring costs, which is of great significance for the government regulation and public prevention of the air pollution.

K-anonymity model for privacy-preserving soccer fitness data publishing

With the development of data mining technology, more and more researchers use the soccer fitness data to analyse the ranking of soccer athletes' and professional training. However, the direct release of soccer fitness data may leak the personal privacy of soccer athletes, so how to ensure the utility of soccer fitness data and the privacy of soccer player has become an issue. In this paper, we point out the linking attack existing in soccer fitness data, which the attackers can use the auxiliary demographic data as background information to attack the published physical data. So the attackers will map the privacy data and the athlete together. At the same time, we apply the partitioning-based and k-means clustering-based two k-anonymity algorithms to the soccer fitness data publishing to trade-offs the data utility and the personal privacy. Experimental results showed that the performance of methods is convincing.

MOOC Education Mode of PE Public Curriculum in Colleges and Universities Based on Data Mining

With the in-depth development of data mining technology and quality education, the teaching of PE public curriculum in colleges and universities has become an important component in the comprehensive cultivation of students’ ability and quality. However, there are many problems in the PE public curriculum in colleges and universities in China. For example, the teaching contents are outdated; the physical education teaching method is single; and it is difficult to satisfy the individualized needs of students. As the most popular courses on the Internet in recent years, MOOC has broken the fixed study time and place so that students can arrange the learning content and plan according to their own interests. This can truly realize the personalized education and open up the new mode of PE public curriculum in colleges and universities. Therefore, the paper takes data mining technology and MOOC teaching mode as the basis of theoretical research, makes a reasonable evaluation of the traditional teaching mode of colleges and universities and then establishes the flipped classroom and hybrid teaching mode of PE public curriculum in colleges and universities under the MOOC teaching mode. Finally, the specific implementation of the MOOC teaching mode is given.

Condition monitoring and fault diagnosis of motor bearings using undersampled vibration signals from a wireless sensor network

Wireless sensor networks (WSNs) which consist of miscellaneous sensors are used frequently in monitoring vital equipment. Benefiting from the development of data mining technologies, the massive data generated by sensors facilitate condition monitoring and fault diagnosis. However, too much data increase storage space, energy consumption, and computing resource, which can be considered fatal weaknesses for a WSN with limited resources. This study investigates a new method for motor bearings condition monitoring and fault diagnosis using the undersampled vibration signals acquired from a WSN. The proposed method, which is a fusion of the kurtogram, analog domain bandpass filtering, bandpass sampling, and demodulated resonance technique, can reduce the sampled data length while retaining the monitoring and diagnosis performance. A WSN prototype was designed, and simulations and experiments were conducted to evaluate the effectiveness and efficiency of the proposed method. Experimental results indicated that the sampled data length and transmission time of the proposed method result in a decrease of over 80% in comparison with that of the traditional method. Therefore, the proposed method indicates potential applications on condition monitoring and fault diagnosis of motor bearings installed in remote areas, such as wind farms and offshore platforms.

Mining High Utility Patterns in One Phase without Generating Candidates

Utility mining is a new development of data mining technology. Among utility mining problems, utility mining with the itemset share framework is a hard one as no anti-monotonicity property holds with the interestingness measure. Prior works on this problem all employ a two-phase, candidate generation approach with one exception that is however inefficient and not scalable with large databases. The two-phase approach suffers from scalability issue due to the huge number of candidates. This paper proposes a novel algorithm that finds high utility patterns in a single phase without generating candidates. The novelties lie in a high utility pattern growth approach, a lookahead strategy, and a linear data structure. Concretely, our pattern growth approach is to search a reverse set enumeration tree and to prune search space by utility upper bounding. We also look ahead to identify high utility patterns without enumeration by a closure property and a singleton property. Our linear data structure enables us to compute a tight bound for powerful pruning and to directly identify high utility patterns in an efficient and scalable way, which targets the root cause with prior algorithms. Extensive experiments on sparse and dense, synthetic and real world data suggest that our algorithm is up to 1 to 3 orders of magnitude more efficient and is more scalable than the state-of-the-art algorithm

Application of Data Mining in Web-based Education Platform

This paper discusses the application of data mining on all aspects of the life-long education platform. Firstly, this paper introduces knowledge and common technology of data mining. Secondly, it introduces the application of data mining technology on course recommendation in the life-long education platform Thirdly, it introduces the application of data mining technology on score analysis in the life-long education platform. Lastly, it introduces the application of data mining technology on assisting teacher’s decisions in the life-long education platform. It is beneficial to introduce data mining technology into web-based education. The data mining technology covers a wide area. This paper will discuss the application of data mining on web-based education from the aforementioned three aspects. With the development of data mining technology and educational information, this paper believes that more and updated technology will be applied to web-based education to make web-based education more intelligent and humanized.

Review on Mining High Utility Patterns Decreasing Candidates

Utility mining is a latest development of data mining technology. Among utility mining issues, utility mining with the item set share structure is a hard one as no anti-monotonicity property hold swith the interestingness measure.Prior work son this problem all use a two-phase,candidate generation approach with one exception that is however in efficient and not scalable with big databases.This paper plan sa novel algorithm that find shigh utility patterns in a single phase with out generating candidate