Mining Methodology Research Articles

Multi-method Coal Seam Floor Water Inrush Risk Evaluation Based on Variable Weight Theory

Frequent water inrush disasters in deep coal seam mining pose a significant threat to the safety of coal extraction operations. Due to the complexity and non-linearity of water inrush factors, evaluating the risk of water inrush in coal seam floor is challenging and the results can vary significantly. To achieve more accurate evaluations, the weighted rank-sum method and Grey-TOPSIS method were employed, alongside a master control indicator variable-weight model, for assessing the risk of water inrush in coal seam floor. Validation of the evaluation zoning map against actual conditions revealed that both methods produced accurate assessment results, thus affirming the reliability of the new evaluation approach. Compared with the water inrush coefficient method, the diversification of index factors weakened the absolute control effect of the water inrush threshold. The evaluation outcomes were more systematic and comprehensive, providing a new methodology and perspective for deep coal seam mining.

Validation of an interactive process mining methodology for clinical epidemiology through a cohort study on chronic kidney disease progression.

Process mining holds promise for analysing longitudinal data in clinical epidemiology, yet its application remains limited. The objective of this study was to propose and evaluate a methodology for applying process mining techniques in observational clinical epidemiology. We propose a methodology that integrates a cohort study design with data-driven process mining, with an eight-step approach, including data collection, data extraction and curation, event-log generation, process discovery, process abstraction, hypothesis generation, statistical testing, and prediction. These steps facilitate the discovery of disease progression patterns. We implemented our proposed methodology in a cohort study comparing new users of proton pump inhibitors (PPI) and H2 blockers (H2B). PPI usage was associated with a higher risk of disease progression compared to H2B usage, including a greater than 30% decline in estimated Glomerular Filtration Rate (eGFR) (Hazard Ratio [HR] 1.6, 95% Confidence Interval [CI] 1.4-1.8), as well as increased all-cause mortality (HR 3.0, 95% CI 2.1-4.4). Furthermore, we investigated the associations between each transition and covariates such as age, gender, and comorbidities, offering deeper insights into disease progression dynamics. Additionally, a risk prediction tool was developed to estimate the transition probability for an individual at a future time. The proposed methodology bridges the gap between process mining and epidemiological studies, providing a useful approach to investigating disease progression and risk factors. The synergy between these fields enhances the depth of study findings and fosters the discovery of new insights and ideas.

Analyzing Daytime/Nighttime Pedestrian Crash Patterns in Michigan Using Unsupervised Machine Learning Techniques and their Potential as a Decision-Making Tool

Background Data mining applications are becoming increasingly common across various fields. Numerous data mining methodologies have been utilized in pedestrian crash data analysis. However, association and correspondence analyses have yet to be extensively employed in pedestrian safety literature to support decision-making. Material and Methods Road lighting significantly affects pedestrian crashes, highlighting the importance of examining pedestrian crash patterns during daytime and nighttime. This study analyzed pedestrian fatal and injury crashes in Michigan from 2011 to 2021 under varying road lighting conditions. The study identified pedestrian crash patterns using unsupervised machine-learning techniques based on several multidimensional factors. The Association Rules Learning (ARL) technique was used to determine general associations of patterns leading to pedestrian crashes. At the same time, Multiple Correspondence Analysis (MCA) helped understand crash attribute patterns in two notable scenarios: elderly pedestrian crashes during daytime lighting and high-speed midblock crashes during nighttime. These methods revealed key patterns associated with contributing factors to pedestrian crashes in different lighting conditions. Results In the analysis of cloud combinations involving elderly pedestrians during daylight, it was found that: 1) elderly pedestrians were significantly involved in severe injury crashes at two-lane midblock locations with a speed limit between 45 and 65 mph; 2) improper actions by elderly drivers were significantly associated with elderly pedestrians on wide roads (>5 lanes); and 3) the complex interaction between city streets with 3-5 lanes was significantly correlated with drivers 'failed to yield' actions. Additionally, in the analysis of cloud combinations involving nighttime pedestrian crashes at high-speed midblock locations, it was found that 1) a specific age group of alcohol- or drug-involved pedestrians (19-30) was significantly associated with young drivers and improper driver actions; 2) during the winter season, young pedestrians were significantly involved in moderate injury crashes at undivided midblock locations; and 3) elderly drivers and 'failed to yield' actions were highly correlated with 3-5 lane roadways. Discussion The research findings are expected to raise awareness of Michigan pedestrian crash patterns under daytime and nighttime lighting conditions. They will also recommend safety countermeasures for practitioners to enhance pedestrian safety in walkable cities. The proposed methodologies can uncover relationships that need to be well-documented in the existing literature on pedestrian safety. Additionally, these methods can identify valuable relationships without restricting variables to being either dependent or independent, and they can reveal relationships that might be challenging to detect otherwise. The results will also provide decision rules and visualizations that are easy to understand. Conclusion The results showed that the proposed techniques can analyze pedestrian crash data in a way that aligns with understanding pedestrian crash characteristics. Traffic safety administrators could benefit from this methodology as a decision-support tool.

Analyzing assisted reproductive treatment representations in Italy and Spain through newspapers.

With the increase in infertility cases recorded over the last 20 years, there is a considerable demand for assisted reproductive treatments (ART). However, there is significant variation in the availability of such treatments across different countries. Legislation on assisted reproduction is influenced by various cultural expressions, making it seemingly impossible to create a single representation adaptable to different contexts. This work investigates the cultural representations of ART in Italy and Spain. We collected 1,735 articles from two Italian and Spanish newspapers, with all the articles containing the respective translation of ART from 2013 to 2022. The two corpora were analyzed using the Emotional Text Mining (ETM) methodology. The analysis produced 3 clusters in the Italian corpus and 5 clusters in the Spanish corpus. From the Italian results, a view of ART emerged that is linked to ethical limitations and the ideal of the traditional family. In contrast, the Spanish results depict ART in terms of community, rights, public health, and birth seen in itself. In conclusion, this study highlights the strong differences between Italian and Spanish cultures regarding ART. The results could be used to improve clinical practices and legislation surrounding ART.

Foundational Segmentation Models and Clinical Data Mining Enable Accurate Computer Vision for Lung Cancer.

This study aims to assess the effectiveness of integrating Segment Anything Model (SAM) and its variant MedSAM into the automated mining, object detection, and segmentation (MODS) methodology for developing robust lung cancer detection and segmentation models without post hoc labeling of training images. In a retrospective analysis, 10,000 chest computed tomography scans from patients with lung cancer were mined. Line measurement annotations were converted to bounding boxes, excluding boxes < 1cm or > 7cm. The You Only Look Once object detection architecture was used for teacher-student learning to label unannotated lesions on the training images. Subsequently, a final tumor detection model was trained and employed with SAM and MedSAM for tumor segmentation. Model performance was assessed on a manually annotated test dataset, with additional evaluations conducted on an external lung cancer dataset before and after detection model fine-tuning. Bootstrap resampling was used to calculate 95% confidence intervals. Data mining yielded 10,789 line annotations, resulting in 5403 training boxes. The baseline detection model achieved an internal F1 score of 0.847, improving to 0.860 after self-labeling. Tumor segmentation using the final detection model attained internal Dice similarity coefficients (DSCs) of 0.842 (SAM) and 0.822 (MedSAM). After fine-tuning, external validation showed an F1 of 0.832 and DSCs of 0.802 (SAM) and 0.804 (MedSAM). Integrating foundational segmentation models into the MODS framework results in high-performing lung cancer detection and segmentation models using only mined clinical data. Both SAM and MedSAM hold promise as foundational segmentation models for radiology images.

Analyzing Chinese Tourists’ Experiences in Seoul and Tokyo: A Mixed Methodology of Text Mining and Interviews

Analyzing Chinese Tourists’ Experiences in Seoul and Tokyo: A Mixed Methodology of Text Mining and Interviews

Public Behavior and Emotion Correlation Mining Driven by Aspect From News Corpus.

Emotion motivates behavior. Investigating the correlation between behavior and emotion, an often overlooked perspective, plays a significant role in uncovering the underlying motives behind behaviors and the intrinsic cause-effects of social events. This article proposes a methodology for mining the correlation between public behavior and emotion using daily news data. Initially, aspect-emotion-reaction (A-E-R) triplets are extracted and generalized, encompassing both explicit and implicit patterns. Then, a knowledge representation model based on hypothetical context (KRHC) with a self-reflection mechanism is proposed to uncover implicit relationships between emotion and behavior through attention mechanisms. By combining rule-based methods for explicit relationships and deep learning for implicit ones, an understanding of emotion-behavior patterns is achieved. In this study, the behaviors are divided into three categories of prosocial, antisocial, and normal behaviors with ten secondary types. Seven categories of emotions are adopted. The proposed deep learning model KRHC is validated on A-E-R datasets and public KINSHIP datasets. The experiment results are concluded; for example, when "fear", "sad", and "surprise" emotions appear, it drives behavior "panic" with most probability. These findings could provide insights for both human-computer interaction and public safety management applications.

Computer vision solution for uniform adherence in gastronomy schools: An artificial intelligence case study

This research study presents an innovative application of computer vision technology in culinary education to ensure consistent student uniform adherence, crucial to accomplishing hygiene, safety, and professionalism standards. The proposed approach utilizes the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology to generate a computer vision application prototype to identify specific culinary uniforms components, such as chef's jackets, aprons, hats, and pants. The development process using LandingLens, a code-free platform, involves several stages: business and data understanding, preparation, modeling, evaluation, and deployment. The final model was training with 77 images, and the application deployment was tested using 38 images. Results demonstrate the potential of artificial intelligence to enhance operational efficiency and uphold professional standards in culinary education. Integrating computer vision addresses the challenges associated with manual monitoring and opens opportunities for broader adoption of technology in culinary pedagogy and training.

Using Machine Learning Models to Detect the Increasing Threats of Financial Fraud in the Cyberspace

In the dynamic landscape of the financial sector, the escalating menace of financial fraud presents pervasive implications for businesses and consumers alike. Particularly, detecting credit card fraud in real- time transactions has become a pivotal concern within the financial industry. This abstract delves into the critical role of data mining in addressing the complexities of credit card fraud detection, shedding light on the multifaceted challenges that confront this domain. The realm of financial business is increasingly besieged by the spectre of financial fraud, necessitating robust measures to combat its detrimental effects. As the sophistication and prevalence of fraudulent activities continue to evolve, the imperative of deploying effective strategies for fraud detection becomes more pronounced. Applying data mining techniques in this context is paramount in identifying and mitigating credit card fraud. Leveraging advanced data mining methodologies is essential for scrutinising live transactions and discerning anomalous patterns indicative of fraudulent behaviour. Credit card fraud detection poses formidable challenges, primarily attributable to two compelling factors. Firstly, the inherent dynamism of normal and fraudulent behavioural profiles engenders a perpetual need for adaptive and responsive detection mechanisms. Secondly, the highly imbalanced nature of credit card fraud data sets further complicates accurately identifying fraudulent activities, necessitating nuanced approaches to discern anomalies amidst voluminous transactional data effectively. In light of the foregoing, this abstract underscore the criticality of data mining in addressing the intricate landscape of credit card fraud detection, emphasising the need for agile and sophisticated methodologies to navigate the evolving nature of fraudulent behaviours and the skewed distribution of fraud-related data sets. By comprehensively elucidating these challenges, this abstract provides a foundational understanding of the nuanced complexities inherent in combatting financial fraud through the lens of data mining.

Causal V: A five-phase causal mining methodology for real-world accident data

Causal V: A five-phase causal mining methodology for real-world accident data

A Survey on Malware Detection and Analysis

Malware, or malicious software, poses a significant threat to the security and functionality of computer systems globally. This survey provides a comprehensive analysis of current malware detection and analysis methods, focusing on data mining methodologies. The study categorizes malware detection techniques into signature-based and behaviour-based approaches, highlighting their respective strengths and weaknesses. It explores heuristic techniques enhanced by artificial intelligence, including neural networks and genetic algorithms, to improve detection accuracy. The literature review examines host-based and network-based intrusion detection systems, hybrid systems, and virtual machine introspection. The paper also discusses static and dynamic analysis methods, emphasizing the importance of analysing malware in controlled environments. Through detailed examination, this survey aims to present a thorough understanding of contemporary malware detection strategies and their applications, offering insights for future advancements in the field.

An Ensemble Machine Learning and Data Mining Approach to Enhance Stroke Prediction.

Stroke poses a significant health threat, affecting millions annually. Early and precise prediction is crucial to providing effective preventive healthcare interventions. This study applied an ensemble machine learning and data mining approach to enhance the effectiveness of stroke prediction. By employing the cross-industry standard process for data mining (CRISP-DM) methodology, various techniques, including random forest, ExtraTrees, XGBoost, artificial neural network (ANN), and genetic algorithm with ANN (GANN) were applied on two benchmark datasets to predict stroke based on several parameters, such as gender, age, various diseases, smoking status, BMI, HighCol, physical activity, hypertension, heart disease, lifestyle, and others. Due to dataset imbalance, Synthetic Minority Oversampling Technique (SMOTE) was applied to the datasets. Hyperparameter tuning optimized the models via grid search and randomized search cross-validation. The evaluation metrics included accuracy, precision, recall, F1-score, and area under the curve (AUC). The experimental results show that the ensemble ExtraTrees classifier achieved the highest accuracy (98.24%) and AUC (98.24%). Random forest also performed well, achieving 98.03% in both accuracy and AUC. Comparisons with state-of-the-art stroke prediction methods revealed that the proposed approach demonstrates superior performance, indicating its potential as a promising method for stroke prediction and offering substantial benefits to healthcare.

A Review on Utilizing Data Mining Techniques for Chronic Kidney Disease Detection

This comprehensive study delves into the application of machine learning (ML) and data mining techniques for the prognosis and diagnosis of Chronic Kidney Disease (CKD), a significant global health concern characterized by the gradual loss of kidney function. Through a detailed examination of various predictive models, the research evaluates the efficacy of different ML algorithms and data mining methodologies in classifying and diagnosing CKD. Utilizing datasets from the UCI machine learning repository and other sources, this study explores a range of ML algorithms-including logistic regression, decision trees, support vector machines, random forest, and deep learning networks-alongside feature selection techniques to enhance prediction accuracy and facilitate early diagnosis. Despite facing challenges such as dataset limitations and the need for external validation, the findings reveal remarkable potential in using ML and data mining to improve CKD diagnosis, with some models achieving accuracy rates exceeding 99%. The research underscores the critical role of technology in advancing CKD diagnosis and management, paving the way for more personalized and effective healthcare solutions.

Analysis of the Spread and Evolution of COVID-19 Mutations in Ecuador Using Open Data.

Currently, the analyses of and prediction using COVID-19-related data extracted from patient information repositories compiled by hospitals and health organizations are of paramount importance. These efforts significantly contribute to vaccine development and the formulation of contingency techniques, providing essential tools to prevent resurgence and to effectively manage the spread of the disease. In this context, the present research focuses on analyzing the biological information of the SARS-CoV-2 viral gene sequences and the clinical data of COVID-19-affected patients using publicly accessible data from Ecuador. This involves considering variables such as age, gender, and geographical location to understand the evolution of mutations and their distributions across Ecuadorian provinces. The Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology is applied for data analysis. Various data preprocessing and statistical analysis techniques are employed, including Pearson correlation, the chi-square test, and analysis of variance (ANOVA). Statistical diagrams and charts are used to facilitate a better visualization of the results. The results illuminate the genetic diversity of the virus and its correlation with clinical variables, offering a comprehensive understanding of the dynamics of COVID-19 spread in Ecuador. Critical variables influencing population vulnerability are highlighted, and the findings underscore the significance of mutation monitoring and indicate a need for global expansion of the research area.

Proof-Theoretical Aspects of Nonlinear and Set-Valued Analysis

AbstractThis thesis is concerned with extending the underlying logical approach as well as the breadth of applications of the proof mining program to various (mostly previously untreated) areas of nonlinear analysis and optimization, with a particular focus being placed on topics which involve set-valued operators.For this, we extend the current logical methodology of proof mining by new systems and corresponding so-called logical metatheorems that cover these more involved areas of nonlinear analysis. Most of these systems crucially rely on the use of intensional methods, treating sets with potentially high quantifier complexity in the defining matrix via characteristic functions and axioms that describe only their properties and do not completely characterize the elements of the sets.The applicability of all of these metatheorems is then substantiated by a range of case studies for the respective areas which in particular also highlight the naturalness of the use of intensional methods in the design of the corresponding systems.The first new area covered thereby is the theory of nonlinear semigroups induced by corresponding evolution equations for accretive operators. In that context, we present (besides an initial foray into the area from 2015) essentially the first applications of proof mining to the theory of partial differential equations. Concretely, we provide quantitative versions of four central results on the asymptotic behavior of solutions to such equations.The second new area unlocked in this thesis is that of the continuous dual of a Banach space and its norm (which are also approached via intensional methods). This in particular relies on a proof-theoretically tame treatment of suprema over (certain) bounded sets in this intensional context which is further exploited later on. These systems, which give access to this until now untreated fundamental notion from functional analysis, are then used to provide further substantial extensions to treat various notions from convex analysis like the Fréchet derivative of a convex function, Fenchel conjugates, Bregman distances, and monotone operators on Banach spaces in the sense of Browder.These systems are then utilized to provide applications in the context of Picard- and Halpern-style iterations of so-called Bregman strongly nonexpansive mappings where we provide both new quantitative and qualitative results.Lastly, we discuss the key notion of extensionality of a set-valued operator and its relation to set-theoretic maximality principles in more depth (which was already singled out—to some degree—in previous work). We thereby exhibit an issue arising with treating full extensionality in the context of these intensional approaches to set-valued operators and present useful fragments of the full extensionality statement where these issues are avoided.Corresponding to these fragments, we discuss a range of uniform continuity statements for set-valued operators beyond the usual notion involving the Hausdorff-metric. In particular, in that context, we utilize the previous tame treatment of suprema over bounded sets to also provide the first proof-theoretic treatment of that Hausdorff-metric in the context of systems for proof mining.The applicability of this treatment of the Hausdorff-metric is then in particular substantiated by a last case study where we provide quantitative information for a Mann-type iteration of set-valued mappings which are nonexpansive w.r.t. the Hausdorff-metric.The abstract was taken directly from the thesis.Abstract prepared by Nicholas Pischke, Technische Universität Darmstadt, Darmstadt, GermanyE-mail: pischke@mathematik.tu-darmstadt.de.URL: https://doi.org/10.26083/tuprints-00026584.

An adaptive methodology based on predictive deep learning and context aware clustering for electricity power usage mining and optimization at different granularity levels

Smart metering in electricity power grid is an optimistic trend at global level. All smart devices and appliances based on Internet of Things (IoT) are now playing very significant role in household. These days’ electric power usage mining and optimization is possible down to meter level only. However, it is very challenging and significant to go down to different granularity levels such as appliances, various sensors and activities etc. The shifting of the electric power usage to low price electricity is also significant and possible by mining and optimizing electric power usage behaviour at low level. All smart appliances and activities are needs to be customized to when you use them. This paper proposes an adaptive methodology based on predictive deep learning and context aware clustering to discover new ways for mining and optimization of electric power usage at different granularity levels and make optimal decisions for shifting electric power usage to low cost. Here we have considered households and business meters approximately 2000 with unique id of each meter. The data of three months is used for user preference of starting appliance. The predictive accuracy of proposed methodology for usage mining and optimization is improved by average 4 %. Different input data features are used to form clusters of meters with similar power consumption behaviour for household occupancy. The clustering accuracy for household occupancy is improved from 0.68 to 0.91. The impact of accurate household occupancy detection and appliance usage mining and optimization is in reduction of electric power cost. The consumer can see how electric power efficiency and time-of-use shift makes a difference using experimental setup.

Machine Learning-Based Cardiovascular Disease Detection and Intelligent Healthcare Workflow Automation

Cardiovascular diseases (CVDs) are a major cause of death worldwide, emphasizing the importance of better early detection methods. This study introduces a machine learning model to detect CVDs by analyzing vital sign information. The model, which was trained using a cardiovascular dataset, utilizes three algorithms: Support Vector Machine (SVM), Naïve Bayes, and Decision Tree. It assesses the health status of patients by predicting vital sign values, allowing healthcare providers to receive timely alerts. Following the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology, a proof-of-concept web application was created with object-oriented design principles and implemented using Python. This application predicts the likelihood of CVD and streamlines the scheduling of doctor appointments, promoting quick medical intervention. Our findings reveal that the Decision Tree classifier performed the best in accurately identifying patients who are at risk based on abnormalities in vital signs. This method can potentially enhance early detection of CVDs and improve the timing of medical care.

A Review on Classification Algorithm for Customer Churn Classification

Any sector faces a huge obstacle when it comes to retaining existing customers. The percentage of consumers who have quit using a product or service is referred to as customer churn, and it is a vital indication that offers reliable information about this percentage. When it comes to achieving long-term success in a market or industry, one of the most significant challenges that any company must face is the ability to keep their precious clients and to fulfill their needs. A review of the most significant studies on Customer Churn Prediction is presented in this paper so as to furnish the reader with an overview of frequently employed data mining methodologies and their respective performances. We provide the available statistics in addition to customer information in order to approximate customer attrition. The time period encompassing the survey extends from 2003 to 2023. During the process of Customer Churn Prediction, we identified the issues and difficulties that were linked with it and offered guidance and potential remedies.

Providing a hybrid approach to increase the accuracy of intrusion detection systems in computer networks

Intrusion detection is a critical obstacle in the realm of security and data mining methodologies. Consequently, researchers have extensively investigated the quest for the swiftest and most precise means of identifying intrusions. Essentially, intrusion detection systems are tasked with recognizing any unauthorized activities, misuse, or harm inflicted upon a system, be it by internal or external users. Recently, in order to design intrusion detection systems, artificial intelligence and machine learning methods have been used, each of which has its own characteristics and advantages. Accordingly, this article focuses on using machine learning to improve the accuracy of the intrusion detection process. In fact, by using machine learning, trends, and patterns can be easily identified and thus used in a network environment to detect intrusion. It can be very useful. For this purpose, we utilize Radial Basis Function (RBF) neural networks and support vector machine (SVM) algorithm to improve decision-making and intrusion detection. By employing RBF neural networks, important features of the data are extracted, which in turnPlease check if the author details and affiliations are presented correctly. Kindly amend if necessary. enhance the overall performance of the solution and the efficiency of the SVM algorithm. This is because feature reduction ultimately leads to improved effectiveness of the SVM algorithm. In methods lacking this capability, the learning algorithm is compelled to utilize features that have no specific correlation with intrusion and essentially do not contribute to identifying attacks. Such learning approaches essentially learn from noisy data, which negatively impacts the intrusion detection solution. Finally, the proposed solution was evaluated using Python programming language and KDD99 data set. The results of the evaluation indicate that the proposed solution has a higher accuracy and precision than other evaluated solutions. So, the accuracy is 97, and the precision is over 99%.

TEXT BASED MASS OPINION MINING USING NEURAL NETWORK ALGORITHM WITH ABSTRACTIVE SUMMARIZATION

Our work, which focuses specifically on Twitter data, presents an advanced methodology for large-scale opinion mining. Sentiment analysis, or opinion mining, is the process of mechanically identifying and categorizing sentiments from textual data. We achieve accurate sentiment categorization and informative result summarizing by combining the latest neural network methods with abstractive summarization techniques. Tokenization, a crucial stage in natural language processing (NLP) activities, is where we start by using BERT (Bidirectional Encoder Representations from Transformers). By capturing the contextual subtleties of language used in Twitter tweets, BERT's contextual embeddings facilitate accurate tokenization. Our model successfully captures the textual data by utilizing BERT's capabilities, which paves the way for further sentiment analysis. We then apply Bidirectional Long Short-Term Memory (BiLSTM) to sentiment categorization. Recurrent neural networks (RNNs) such as BiLSTM are particularly good at recognizing sequential dependencies in data. Sentiment analysis relies heavily on word order since sentiments are frequently influenced by the context that words that come before and after give. Our model's incorporation of BiLSTM allows it to precisely classify tweets as either positive or negative by capturing these sequential dependencies. We also present an abstractive summary element to produce brief summaries that capture dominant attitudes in the dataset. Abstractive summarization generates logical summaries that encapsulate the main ideas of numerous tweets, going beyond simple sentence selection and concatenation. We evaluate the quality of the generated summaries and the accuracy of sentiment categorization achieved by our model through extensive experimentation on a representative and diversified Twitter dataset. Our findings show how reliable and effective our approach is at identifying significant sentiment trends and offering insightful information about the general thoughts shared on Twitter. In the end, our work advances the field of opinion mining and provides useful instruments for deciphering and evaluating massive amounts of social media data.