Journal of Data Mining and Knowledge Engineering ISSN: 3139-5341 (Online)

Abstract

A smart world, such as ours, is essentially built on the Internet of Things (IoT). Every day, the IoT creates massive amounts of data. It is critical to work on this generated data in order to make appropriate use of it. This is mostly accomplished through the use of data mining. This research examines how data mining may be used on IoT data. Classification (grouping data), clustering (labelling data), frequent pattern mining (identifying often occurring itemsets, sequences, or substructures in data), and outlier analysis are the main components of data mining (analyzing data with abnormal values of attributes). This paper's major focus is on frequent pattern mining.

Keywords: Frequent Pattern Mining, Internet of Things, Data Mining

Abstract

Disease prediction in the healthcare business is greatly aided by data mining. Data mining techniques are often employed in early disease diagnosis. Diabetes is one of the world's most serious health issues. Diabetes is a common chronic illness. Diabetes prediction is a rapidly expanding science. Diabetes prediction in an early stage will lead to more effective treatment. Diabetes is vital to avoid, monitor, and raise diabetes awareness since it causes other health problems. Diabetes, whether type 1 or type 2, can cause heart problems, renal disease, and eye difficulties. This comprehensive study examines several techniques and data mining methodologies for predicting diverse diabetic problems at an early stage. Diabetes has caused it to become a chronic illness. Early detection of this condition will save patients' lives. Diabetes is averted and treatment rates are decreased by the use of data mining tools and methods. Among the most frequent and essential data mining technologies are association rule mining, classification, clustering, Random Forest, prediction, and the Artificial Neural Network (ANN). There are several data mining approaches accessible to avoid ailments such as heart disease, cancer, and renal disease, among others. This study looks at how data mining tools may be used to forecast numerous kinds of illnesses.

Keywords: Data Mining, Diabetes Diseases, Prediction of Diabetes, Data Mining Techniques

Abstract

Over the years, the most popular issue has been the mining of common objects from large amounts of data storage. Market basket analysis is one of the many real-world applications of frequent pattern mining. We offer an overview of recent frequent pattern mining strategies employing data mining algorithms in this study. Frequent pattern mining in data mining necessitates a large number of database scans. As a result, it is a computationally costly process. As a result, there is still a need to update and improve existing frequent pattern mining approaches in order to obtain more efficient solutions for the same activity. A review of all the contemporary and most popular frequent pattern mining techniques is also undertaken in this work.

Keywords: Data Mining, Frequent Item Mining, Support, Confidence, Market Basket Analysis, Parallel Execution

Abstract

As an underlying paradigm for product development management, a hybrid knowledge model is proposed. It is capable of supporting hybrid features such as ontologies and rules. Effective product development cooperation is dependent on the sharing and reasoning of product information as well as engineering knowledge. Much research has taken product information and engineering expertise into account. However, much past research has concentrated on either developing ontology of product information or developing rule-based engineering knowledge systems. This research demonstrates how a hybrid F-logic-based knowledge model might enable such desirable characteristics.

Keywords: Product development, Ontology, Rule, and F-logic

Abstract

Privacy-preserving data mining is a critical topic in data mining today. Because numerous companies and individuals are producing sensitive data or information these days, they do not wish to divulge their sensitive data, yet that data might be beneficial for data mining. As a result of privacy preserving mining, such data may be mined usefully without jeopardizing its privacy. Because the data is now safe owing to encryption, privacy may be retained by encrypting the database that is to be mined. Code profiling is a branch of software engineering in which we may utilize data mining to identify knowledge that will be valuable in future software development. We used privacy-preserving mining of code profiling data such as software metrics of diverse codes in this work. The accuracy of data mining results on real and encrypted data is compared. We also looked at the outcomes of privacy preserving mining in code profiling data and discovered some intriguing results.

Keywords: Data mining, Privacy protection, code profiling, and correlation coefficient

Abstract

The system proposed here contains a variety of medical tools such as sensors and web-based applications or applications that connect to network devices and help monitor and record patient health data and medical information. Remote health caring of patient reception is increasing with the recognition of assorted nature of mobile devices that has developed to enable remotely caring. The cloud also as IoT (Internet of Things) and therefore the mobile technologies make it easier to observe the patient health conditions by sharing the health information to health care teams like doctors, nurses and specialists. However, the guardians of the patients are often anxious about their patients after they are in work. By ensuring guidance awareness about the patients, it can bring more liability of the hospital management. We’ve got demonstrated a health care system for hospital management to permit guardians together with doctors to remotely monitor health conditions of patients via internet. Remote monitoring and guidance awareness by sharing information in a much authenticated manner are the most focus.

Keywords: Internet of Things, IoT in Healthcare, Patient Monitoring, Arduino uno, Smart Health Monitoring.

Abstract

India is planning an integral direction of startup welcoming policies and a business-friendly or entrepreneur’s environment. India is one of the fastest growing countries in terms of entrepreneurship. Entrepreneurship is an essential input for economic development, creating new markets or opportunities. India is now pushing entrepreneurship passionately, but creating jobs through startups is a significant challenging for both the government and the Indian public. Startup India is a flagship initiative of the Indian government aimed at fostering entrepreneurship and innovation. The government intends to empower startup companies through this effort in order to increase entrepreneurship, economic growth, and employment across India. The startup ecosystem in India has become a global topic of discussion. This paper intention at offering an analytical overview of the boom and potentialities of startup systems in India i.e. the progress made by India so far. Therefore, this research can contribute to a better understanding of the Investment and financing strategy of entrepreneurial ventures.

Keywords: Startups, Entrepreneurs, Analytical, Business-friendly

Abstract

Crime is an act harmful not only to some individual but also to a community, society, or the state. Such acts are forbidden and punishable by law. Crimes are a global problem as they can affect economic growth and quality of life. It can also harm the nation under both social and economic conditions. Crime control is an inescapable step that is compulsory for the welfare and sustainable development of a nation. We know very well that in the digital world, it is not an easy task to expose the criminals and the vulnerable areas that are continuously getting affected by their wrongdoings. The police departments of every nation are also continuously working in a paced manner to overcome the crimes, criminals and their techniques. The difficulty in investigating a large amount of data regarding crimes and criminals has become a major challenge for police department officials. An approach is needed that can classify, systematically investigate and forecast the crimes that help to reduce the crime rate. There are various methodologies and paradigms which will help police officials to discover and eliminate crimes from society. Data mining empowers us with several practical and convenient ways to assess large and distinct sets of information. It helps to uncover hidden information from the large database of criminal records for investigating, controlling and preventing crime for organizations and users. Various researchers and data analysts gave their valuable time and knowledge to the field of data mining. The paper focuses primarily on presenting a short and snappy overview of various research papers focused on the techniques of data mining that have been applied in crime analysis. The dataset used in this paper is taken from the Kaggle. The analysis part identifies the number of unsolved crimes. The parameters used are analysis of the various factors like the year, month, location, gender, type of crime used in the unsolved crimes.

Keywords: Data mining, Crime analysis, Criminal investigation, Machine Learning, KNN, Clustering

Abstract

Data mining has recently grown in favour among researchers. Data mining offers a variety of strategies and approaches for analysing data generated by various applications in various domains. Similarly, educational mining offers a method for examining educational data sets. Educational mining is concerned with establishing ways for uncovering information from educational data, and it aids in the extraction of hidden patterns and the discovery of new knowledge from vast educational datasets via the use of data mining techniques and tools. Educational mining knowledge may be utilised to make decisions at higher education institutions. This work is based on a survey of the literature on various data mining approaches, as well as specific algorithms such as classification, clustering, and so on. This research examines the efficacy of mining approaches with an educational data set for higher education institutions.

Keywords: Data mining; Higher Education Data, Classification; Educational Mining; Clustering

Abstract

Data mining is a technique for extracting meaningful information from vast amounts of data. It's utilised to carry out and solve many sorts of research projects. Text mining, web mining, picture mining, sequential pattern mining, spatial mining, medical mining, multimedia mining, structure mining, and graph mining are some of the study fields in data mining. Text mining is also known as knowledge discovery in the text (KDT) or knowledge of intelligent text analysis, and it is often referred to as text of data mining. The technique transforms unstructured data into semi-structured data, resulting in high quality information. Text mining is the process of mechanically extracting information from a variety of textual sources and extracting new, previously unknown information using a computer. The method of text mining and approaches, tools, applications, and strategies are discussed in this study.

Keywords: Data mining, Knowledge discovery, Information retrieval, Computer, Text mining

Abstract

The employment of digital applications around the world has resulted in the collecting of massive amounts of information. The valuable information buried inside huge amounts of data is drawing scholars from a wide range of fields to conduct research into creating efficient techniques to extracting the hidden knowledge contained within them. When it comes to data mining, the process of extracting or mining meaningful and valuable knowledge from enormous volumes of data may be defined as follows: When it comes to data mining, there are many different domains to choose from, including text mining, picture mining, sequential pattern mining, web mining, and others. Sequence mining is one of the most significant study areas in this field, since it aids in the discovery of sequential relationships that have been discovered in data. It is used in a broad variety of applications, including the study of consumer purchase trends, online access patterns, weather observations, protein sequencing, DNA sequencing, and other types of data mining. For sequence alignment, sequence searching, and sequence classification in protein and DNA studies, sequence mining techniques are employed in conjunction with other approaches. Researchers in the field of protein sequence analysis are demonstrating an increased interest in the field of protein sequence categorization, according to their findings. It has the capability of identifying recurrent structures in protein sequences that are not previously known. This study describes the many methodologies employed by different researchers in the classification of proteins, as well as providing an overview of the numerous protein sequence classification methods available today.

Keywords: Data mining, Information, DNA, Protein sequence, Sequence mining

Abstract

The main objective of the spatial data mining is to discover hidden complex knowledge from spatial and not spatial data despite of their huge amount and the complexity of spatial relationships computing. This paper reviews the data mining methods that are combined with Geographic Information Systems (GIS) for carrying out spatial analysis of geographic data. We will first look at data mining functions as applied to such data and then highlight their specificity compared with their application to classical data. The collected data far exceeded human's ability to analyze. Recent studies on data mining have extended the scope of data mining from relational and transactional databases to spatial databases. We will conclude by discussing the main differences between these two approaches and the elements they have in common. We will further explain the research conducted in this sector and point out that two separate methods exist: one is focused on space database learning while the other is based on space statistics. Finally, we will address the key distinctions between these two methods and their similar features.

Keywords: Spatial Data Mining, Spatial Databases, Rules Induction, Spatial Statistics, Spatial Neighborhood, knowledge discovery

Abstract

Characteristics of the technical and economic data of mining enterprises are multi-dimensionality and nonlinearity. The sales price data of mineral products is an important economic indicator of mining enterprises, and the geological data is an important technical data. The analysis method of the technical and economic data is researched using technologies of big data analysis and data mining. The fluctuation pattern and influencing factors of the mineral products price are analyzed. The prediction model of the mineral products price is established using artificial neural network. The results show that the practicability of the prediction model is strong, and the prediction accuracy is high. During the process of mineral development, due to the limitation of technical conditions and equipment conditions, lots of geological data have been lost, which reduces the accuracy of the ore body shape and that of reserves estimation. Based on techniques of geostatistics and artificial neural network, the prediction model of the geological missing data is established. By using the model, the regularity of geological data of single borehole, the regularity of geological data of group boreholes and the regularity of geological data of all boreholes is discussed and analyzed. It has been proved that most of the geological missing data can be predicted and interpolated, and results of prediction and interpolation are reliable.

Keywords: Mining enterprises; Technical and economic data; BP neural network; prediction models

Abstract

As one of the most prosperous applications of text analysis and understanding emotions and short messaging text apperception has recently received consequential attention especially during the past several years. This is corroborated by the blogs, chats, online reviews, discussions and systematic evaluations of text analysis techniques. There are 5 major aspects for this trend, first is the commercial and social marketing applications, second is the understanding of one‘s feelings, third human-computer interaction , fourth text to speech generation and fifth is the natural language and machine learning approaches and technologies after 20 years of research. This model includes happy or joy, sad, disgust, anger, fear and surprise as basic emotion. In this paper emotion detection using semantic approaches and rule based approaches are described. Keywords: NLP (Natural Language Processing),

Keyword based detection, Text Processing, Sentimental measures, Semantic approach, Parts of Speech (POS).

Abstract

Microarray technology is one of the most challenging areas where researchers study genetics and organisms. Microarray data are a subset of biomedical data that have huge dimensions. The real challenge that such microarray data pose is its enormous features with a very low sample space. This paper presents various pre-processing techniques on microarray data for Feature Selection and Classification. Several microarray datasets were taken for experimental analysis, where both binary and multi-class classification was performed. Extensively large numbers of features were successfully reduced while achieving good classification accuracy.

Keywords: - Microarray data, data pre-processing, Principal Component Analysis (PCA), Analysis of Variance (ANOVA), Recursive feature elimination (RFE), Feature Selection, Dimension Reduction, Machine Learning Algorithms.

Abstract

Children with Autism Spectrum Disorders (ASD) have a deficiency in figuring out conceptual comparable traits which can be tougher to understand, along with emotions and additionally spotting the similarities, which can be tough for getting to know new situations. People with this syndrome have issues with emotion recognition. Our proposed method works on detecting the feelings of autistic kids from the expression on their faces. Facial Emotion Detection is a technique closer to detecting human feelings via facial expressions. The technique additionally contributes to the getting to know feature of emotion recognition for autistic kids. Autistic humans have repetitive behaviour. They are not prepared to do social communication. The proposed method will discover the expression of autistic humans and apprehend the emotion, and offers the description as Text. The challenge consists of the feelings along with happiness, angry, fear, unhappy and neutrality. In our technique, we can additionally upload a function as Audio with voice for his or her smooth understanding. The set of rules used is Convolution Neural Network (CNN). Our proposed device is a software primarily based totally, and it is able to be easily accessed with the Web camera. Our challenge will even offer as a getting-to-know platform online. The software will assist autistic kids in enhancing their communication skills and adjustments in their behaviour and is beneficial in expressing their feelings.

Keywords: - Autism, Emotion Recognition, Facial Expression Recognition, Image processing, Convolution Neural Network. Training.

Abstract

This paper further enhances the techniques and chronological methods to perform the corresponding manipulation and further prediction analysis. We have acquired a real-time dataset based on the Twitter user's comments sections. The uniqueness of the dataset is that we have extracted only the particular comments which syntheses a particular word based on the product. Then further repetitive extraction is made to complete the dataset. Our dataset has three columns based on the ideology that the particular user or our focused subject has enhanced any detail about the product that we are observing. In this precise dataset, we have taken the subject about the usage of the product by the users that have been manufactured by the companies Google and Apple. Both technology giants have well versed their technology reign in this era, and they are further focused on their upcoming cyber projects, and their products will be more advanced in the future. As they are involved in further optimization in their devices and increasing their specifications. It would be a complex task to accomplish their project without the feedback, pros, and cons of their predecessor projects. They can extract such features from the Twitter API dataset, and they could further enhance their product. They could analyze the drawback, whether the product has reached the market and came out with success all type of this information can be extracted from social media instead of conducting a survey. Such a process would be a more hectic classification, and we cannot predict any accurate results. So we proceed with the social media Data mining Process. 

Keywords: - Twitter Data, Dataset, Data Mining, API, Analyze.

Abstract

Nowadays, the concept of Big Data has acquired more popularity in the field of Education, the IT industry, Business, Healthcare and many more areas. Big data is generated every day by the communication of billions of people using mobiles, computers, GPS devices, sensors and medical devices etc. Today big data creates a huge opportunity for those who can use it efficiently. Detection of previously unknown meaningful patterns and knowledge from a few Gigabytes to Exabyte of data collected from mobile users at the network level or the app-level is termed as MBD. The new wave has provided mobile operators with a huge opportunity to collect enormous data to monitor the technical and transactional aspects of their networks. In this paper, we focus on how the flood of data gathered from heterogeneous Mobile applications such as communication-oriented applications, bank-related applications, commercial applications etc. furthermore, we discuss the impact of these gathered data on human life routine. We investigated the usage and effect of WhatsApp and Facebook by conducting the survey. This article focuses on an integrated representation of this emerging field to bridge multiple disciplines and hopefully to motivate future research.

Keywords: - Big Data, MBD, Mobile Applications, WhatsApp, Facebook

Abstract

Decision Support Systems (DSS) are widely used in organizations to assist managers and professionals in making informed decisions. However, traditional DSS mainly rely on structured databases and statistical models, which often fail to capture semantic relationships among heterogeneous data sources. In recent years, Knowledge Graphs (KGs) have emerged as a powerful paradigm for representing structured and unstructured information in a graph based format. Knowledge graphs connect entities and relationships in a meaningful way, allowing reasoning, inference, and contextual understanding. This paper reviews the role of knowledge graphs in enhancing decision support systems. It discusses the architecture of knowledge graph–enabled DSS, integration techniques, reasoning mechanisms, and real-world applications in healthcare, finance, supply chain, and smart governance. The paper also examines challenges such as scalability, data quality, privacy, and interoperability. Finally, future research directions are presented to improve intelligent decision-making through semantic technologies.

Keywords: Knowledge Graphs, Decision Support Systems, Semantic Reasoning, Data Integration, Intelligent Analytics, Ontology, Graph Databases

Abstract

Financial markets generate huge amount of structured and unstructured data every second. The increasing complexity of global financial systems makes traditional statistical approaches less sufficient for accurate forecasting. Predictive analytics combined with data mining techniques has emerged as an effective solution to analyze financial datasets and generate actionable insights. This paper presents a comprehensive review of predictive analytics methods applied to financial data using data mining approaches. Various techniques such as classification, regression, clustering, neural networks, support vector machines, and ensemble learning are discussed. Applications including stock price prediction, credit risk assessment, fraud detection, and portfolio management are examined. The study also compares different algorithms based on accuracy, interpretability, and computational cost. Challenges such as data volatility, overfitting, and model interpretability are highlighted. The paper concludes that while predictive analytics significantly enhances financial decision making, proper feature engineering and risk management are still essential for reliable forecasting.

Keywords: Predictive Analytics, Financial Data Mining, Stock Market Prediction, Credit Risk Analysis, Machine Learning in Finance, Fraud Detection, Time Series Forecasting, Big Data Analytics.

Abstract:  Ontology-based knowledge representation has emerged as a pivotal approach in the design and development of intelligent systems. By providing a structured and formal way to represent knowledge, ontologies enable machines to reason, infer, and interact with complex datasets in a human-like manner. This paper reviews the principles, methodologies, and applications of ontology-based knowledge representation in intelligent systems. It explores ontology modeling techniques, reasoning mechanisms, semantic interoperability, and integration with artificial intelligence (AI) technologies. Additionally, the paper presents comparative analyses of ontology languages, frameworks, and tools commonly used in real-world applications. The research highlights current challenges, including scalability, dynamic knowledge updates, and ontology alignment, while proposing future directions for more adaptive and context-aware intelligent systems.

Keywords: Ontology, Knowledge Representation, Intelligent Systems, Semantic Web, Reasoning, Ontology Languages

Abstract

The rapid growth of the Internet of Things (IoT) has resulted in massive generation of real-time data from sensors, smart devices, and connected systems. This large scale data presents new opportunities for intelligent decision-making across domains such as healthcare, smart cities, agriculture, industry, and transportation. However, mining IoT data is complex due to its volume, velocity, variety, and veracity. This paper presents a comprehensive review of IoT data mining techniques, architectures, challenges, and applications for intelligent decision making. It discusses preprocessing strategies, machine learning approaches, deep learning models, and edge computing frameworks. Furthermore, the paper compares different mining techniques based on accuracy, scalability, and computational cost. Case studies from smart healthcare and industrial IoT are also examined. The study highlights that integration of big data analytics, artificial intelligence, and edge-cloud frameworks significantly improves real-time decisions. However, issues related to privacy, data security, and scalability still remain open research challenges.

Keywords: Internet of Things (IoT), Data Mining, Intelligent Decision-Making, Machine Learning, Edge Computing, Big Data Analytics, Smart Systems

Abstract

The increasing frequency and intensity of natural and man-made disasters have made disaster management a critical concern for governments and communities worldwide. Geospatial data, which includes satellite imagery, remote sensing data, GPS data, and social media location information, has emerged as a powerful tool in disaster risk reduction, mitigation, and response. Mining geospatial data enables the extraction of meaningful patterns and insights to predict, monitor, and manage disasters effectively. This paper reviews current techniques and applications of geospatial data mining in disaster management, focusing on data sources, preprocessing, mining methodologies, and real-world applications in disaster prediction, risk assessment, and emergency response. The paper also discusses challenges and future directions for integrating advanced analytics and artificial intelligence into geospatial disaster management.

Keywords: Geospatial data, disaster management, data mining, remote sensing, GIS, risk assessment, emergency response, spatial analysis

Abstract

Structured data such as tables, graphs, and relational records constitute the backbone of industrial analytics, healthcare informatics, finance, and engineering monitoring systems. However, learning meaningful representations from structured data remains challenging due to heterogeneity, sparsity, and limited labeled samples. Self-supervised representation learning has recently emerged as a promising paradigm to extract informative latent features without heavy reliance on manual annotations. Instead of labels, the learning objective is derived from intrinsic structure of data such as attribute relations, topology, or contextual dependencies. This paper reviews major self supervised learning (SSL) approaches for structured data including tabular data, graph-structured data, relational databases, and time-series tables. We discuss pretext task design, contrastive and generative frameworks, and hybrid architectures combining neural networks with statistical models. Comparative analysis of methods is provided along with application domains such as anomaly detection, recommender systems, predictive maintenance, and healthcare risk modeling. The study also outlines challenges including feature heterogeneity, missing values, scalability, and evaluation metrics. Finally, emerging directions such as multimodal structured SSL and foundation models for tabular learning are discussed. The review aims to provide a consolidated understanding of self-supervised representation learning for structured datasets and its role in future intelligent data systems.

Keywords: Self-supervised learning, structured data, representation learning, tabular data, graph learning, contrastive learning, pretext tasks

Abstract

Knowledge graphs (KGs) have become fundamental in representing and organizing complex relationships among entities across diverse domains. As the volume and velocity of data increase, the need for dynamic and scalable knowledge graph construction becomes critical. Traditional static KGs fail to capture the evolving nature of real-world knowledge, limiting their applicability in domains such as recommendation systems, biomedical research, and social networks. This paper presents a comprehensive review of scalable methods for constructing dynamic knowledge graphs, highlighting both architecture-level and algorithmic solutions. We explore data ingestion strategies, incremental updating mechanisms, distributed processing frameworks, and emerging machine learning techniques for entity and relation extraction. Moreover, we analyze current challenges including graph sparsity, temporal reasoning, and real-time updates, and discuss evaluation metrics suitable for dynamic KGs. Finally, this work provides insights into future directions for building robust, scalable, and adaptive knowledge graph systems.

Keywords: Knowledge Graphs, Dynamic Graphs, Scalability, Graph Construction, Entity Extraction, Temporal Reasoning, Distributed Computing.