Data-driven Mining Research Articles

ARTIFICIAL INTELLIGENCE AND PRIVACY: AN INVESTIGATION ON HOW ARTIFICIAL INTELLIGENCE-BASED DATA MINING IS AFFECTING THE PRIVACY OF USERS

Artificial intelligence has become an integral part of everyone’s lives in today’s world. Due to the rapid advancements in the field of artificial intelligence, there hasn’t been sufficient mainstream analysis about its cons: especially privacy invasion. This demands focused attention on artificial intelligence (AI) driven data mining, a process which calls ethics and user privacy into question. Ultimately, there must be important compromises made, and in this case it’s between privacy and growth.

A 21st Century Approach in Analysing Health Precautions in London with Machine Learning Driven Data Mining

As in the past, today preventive treatments and health policies constitute an important role in combatting with several diseases, medical phenomenon like pandemies or epidemies. These approaches can prevent several health focused negative consequences in early stages or can give leaders and medical professionals advantage in managing risks associated with health concerns. Usually effective usage of early warning systems, analysis of historical data for exploratory and confirmatory understanding may provide several advantages in this context. In this study a historical data analysis has been applied to understand similar phenomena with the help of machine learning driven data mining. Clustering and classification performances and rules generated by these approaches have also been assessed.

A Biological Data-Driven Mining Technique by Using Hybrid Classifiers With Rough Set

Biological data classification and analysis are significant for living organs. A biological data classification is an approach that classifies the organs into a particular group based on their features and characteristics. The objective of this paper is to establish a hybrid approach with naive Bayes, apriori algorithm, and KNN classifier that generates optimal classification rules for finding biological pattern matching. The authors create combined association rules by using naïve Bayes and apriori approach with a rough set for next sequence prediction. First, the large DNA sequence is reduced by using k-nearest approach. They apply association rules by using naïve Bayes and apriori approach for the next sequence pattern. The hybrid approach provides more accuracy than single classifier for biological sequence prediction. The optimized hybrid process needs less execution time for rule generation for massive biological data analysis. The results established that the hybrid approach generally outperforms the other association rule generation approach.

Data Cleaning in Knowledge Discovery Database-Data Mining (KDD-DM)

Data quality is a main issue in quality information management. Data quality problems occur anywhere in information systems. These problems are solved by Data Cleaning (DC). DC is a process used to determine inaccurate, incomplete or unreasonable data and then improve the quality through correcting of detected errors and omissions. Various process of DC have been discussed in the previous studies, but there is no standard or formalized the DC process. The Domain Driven Data Mining (DDDM) is one of the KDD methodology often used for this purpose. This paper review and emphasize the important of DC in data preparation. The future works was also being highlight.

Intelligent Context Driven Data Mining To Analyse Student Performance in Higher Educational Institutions (HEIs)

Context driven research has been conducted by many. However, very less work has been conducted in constructing a context driven data mining that helps in HEI decision-making. A Student Information System that interacts with Students, Faculties, Student Parents and Management might not have enough information of the background. Context driven data mining is an application intelligent enough to detect and examine the context from different sources and take suitable actions to improve performance and efficiency of decision-making by discovering the hidden factors. This paper recommends a context driven data mining method for understanding student performance from Student Information System in HEIs.

Development of groundwater favourability map using GIS-based driven data mining models: an approach for effective groundwater resource management

The development of groundwater favourability map is an effective tool for the sustainability management of groundwater resources in typical agricultural regions, such as southern Perak Province, Malaysia. Assessing the potentiality and pollution vulnerability of groundwater is a fundamental phase of favourability mapping. A geographic information system (GIS)-based Boolean operator of a spatial analyst module was applied to combine a groundwater potentiality map (GPM) model and a groundwater vulnerability to pollution index (GVPI) map, thereby establishing the favourable zones for drinking water exploration in the investigated area. The area GPM model was evaluated by applying a GIS-based Dempster–Shafer–evidential belief function model. In the evaluation, six geoelectrically determined groundwater potential conditioning factors (i.e. overburden resistivity, overburden thickness, aquifer resistivity, aquifer thickness, aquifer transmissivity and hydraulic conductivity) were synthesized by employing the probability-based algorithms of the model. The generated thematic maps of the seven hydrogeological parameters of the DRASTIC model were considered as pollution potential conditioning factors and were analysed with the developed ordered weighted average–DRASTIC index model algorithms to construct the GVPI map. Approximately 88.8 and 85.71% prediction accuracies for the Groundwater Potentiality and GVPI maps were established using the reacting operating characteristic curve method and water quality status–vulnerability zone relationship scheme, respectively. Finally, the area groundwater favourability map (GFM) model was produced by applying a GIS-based Boolean operator on the Groundwater Potentiality and GVPI maps. The GFM model reveals three distinct zones: ‘not suitable’, ‘less suitable’ and ‘very suitable’ zones. The area analysis of the GFM model indicates that more than 50% of the study area is covered by the ‘very suitable’ zones. Results produce a suitability map that can be used by local authorities for the exploitation and management of drinking water in the area. The study findings can also be applied as a tool to help increase public awareness of groundwater issues in developing countries.

An Improvement of Knowledge Discovery Database (KDD) Framework for Effective Decision

In this study, an understanding and a review of Knowledge Discovery Database (KDD) development and its applications in tire maintenance are highlighted. Even though data mining has been successful in becoming a major component of various business processes and applications, the benefits and real-world expectations are very important to consider. It is also surprising to note that very little is known to date about the usefulness of applying knowledge discovery in transport related research. From the literature, the frameworks for carrying out knowledge discovery and data mining have been revised over the years to meet the business requirements. The Domain Driven Data Mining (DDDM) is one of the KDD frameworks often used for this purpose. In this study, we apply DDDM-KDD for formulating effective tire maintenance strategy within the context of a Malaysian's logistics company. We also discussed the weaknesses of the results from DDDM-KDD and emphasize the important of using the next generation of KDD framework Actionable Knowledge Discovery (AKD) for an effective decision. The direction flow of research, research methods use and contribution of research also are highlighted.

Direct Sum Simplex Signal Space Codes for DNA Nucleotides and Amino Acid Codons

A genomic symbol to signal conversion concept is introduced and demonstrated on nucleotide base pairs and amino acid codons. Binary encodings of the presence or absence of bioinformatic attributes (e.g., physical, chemical, functional, structural, etc.) onto Euclidean signal space points of $+1$ or $-1$ is performed for individual base pairs or blocks of DNA symbols such as amino acid codons. The symbol to Euclidean space mapping is further encoded into an expanded digital representation based on a direct sum $M$ -ary tree structure called a $\sigma$ -tree. For bioinformatic query sequences, the $\sigma$ -tree provides a rapid search mechanism to find near neighbor similarity sets within warehouses of heterogeneous bioinformatic data. Bioinformatic utility will be achieved in data mining systems based on the proposed symbol-to-signal mapping method through trend analysis of bioinformatic attribute data within similarity set aggregates. Query specific feature/attribute similarity sets are revealed by high dimensional Euclidean signal space clusters. The proposed binary encoded signal vectors provide an equal energy Euclidean space simplex code in the sense of channel coding signal constellation design. Use of an equal energy simplex code prevents biases from being introduced among different attributes during the symbol sequence to signal sequence mapping. The inappropriate use of a multienergy, low dimension, signal point mapping is replaced with a same-energy, high dimension, signal point mapping. The $M$ -ary digital encoding and associated near neighbor search method is extensible to nucleotide sequence blocks of any length, producing digitally expressed signal space simplex representations on a hypersphere of high dimension. Moreover, the direct sum structure empowers computation and memory efficiencies that grow only linearly with increasing sequence length and signal set dimensionality. The bioinformatic attribute encoding method and efficient $\sigma$ -tree expanded digital representation introduced here will provide a foundation for constructing novel genomic driven data mining systems.

Semi-Automated Disaggregation of a Conventional Soil Map Using Knowledge Driven Data Mining and Random Forests in the Sonoran Desert, USA

Conventional soil maps (CSM) have provided baseline soil information for land use planning for over 100 years. Although CSMhave been widely used, they are not suitable to meet growing demands for high resolution soil information at field scales. The authors present a repeatable method to disaggregate CSM data into ~30-meter resolution rasterized soil class maps that include continuous representation of probabilistic map uncertainty. Methods include training set creation for original CSM component soil classes from soil-landscape descriptions within the original survey database. Training sets are used to build a random forest predictive model utilizing environmental covariate maps derived from ASTER satellite imagery and the United States Geological Survey (USGS) National Elevation Dataset. Results showed agreement at 70 percent of independent field validation sites and equivalent accuracy between original CSM map units and the disaggregated map. Uncertainty of predictions was mapped by relating prediction frequencies of the random forest model and success rates at validation sites.

COMBINED PATTERN MINING FOR D3M USING FUZZY

Many algorithms and models were developed but the findings are not actionable and lack of soft power while solving the complex problems. Domain Driven Data mining is used a major efforts to promote the action ability of the knowledge discovery in the real world smart decision making. Combined mining is one of the common methods for analyzing complex data for identifying complex knowledge. The deliverables of combined mining are combined patterns. The complex environment gives the combined patterns. In this research we process a new technique called Fuzzy Combined Pattern Mining (FCPM) for Domain Driven Data Mining. It was used to find all the rules that satisfy the minimum support and minimum confidence constraints. In FCPM, we first apply the fuzzy concept to find the patterns after that the fuzzy pattern will be merged to combined pattern mining. The proposed algorithm have been implemented and compared with Apriori Its performance was studied on an experimental basis. The main objective is to provide the interesting patterns to the end user. The implementation of fuzzy in combined mining will generate the rules and based on rules we can identify the interesting patterns.

Semi-automated disaggregation of conventional soil maps using knowledge driven data mining and classification trees

Disaggregation of conventional soil surveys has been identified as a potential source for much of the next generation of model-ready digital soil spatial data. This process aims to apportion vector soil surveys into raster (gridded) representations of the component soils that are often aggregated together in map unit designs. Most soil surveys are published with some description of the soil–landscape relationships that distinguish component soils within map units. We used these descriptions found in the Soil Survey Geographic (SSURGO) database of Webster and Pocahontas Counties in West Virginia, USA, to build a set of representative training areas for all soil components by using 1-arc second digital elevation data and derived geomorphic indices. These training areas were then used in classification tree ensembles with a more extensive environmental database to transform the original SSURGO map into a gridded soil series map. We created underlying prediction frequency surfaces from the models that can be used for creating continuous representations of soil class and property distributions.Disaggregation models matched training sets in 71%–74% of pixels and matched components in original SSURGO map units in 56%–65% of the study area. We evaluated both the original SSURGO data and our models using 87 independent pedons not used in model building. Validation pedons matched components in SSURGO map units at 39% of sites, but in map units that only included one named component (as opposed to multiple soils that could be matched to validation pedons) only 27% of the sites matched. Disaggregation predictions matched validation pedon classes 22–24% of the time using nearest neighbor spatial matches, and these rates increased to 39–44% for correct predictions within a 60-meter radius of the pedon. To characterize uncertainty, we compared relative ensemble prediction frequency (probability) of final hardened model classes at validation sites. Sites with correct predictions had generally higher prediction frequencies; which lead us to use them to create an uncertainty model. Uncertainty was calculated by determining the rate of correct predictions at validation sites within different intervals of prediction frequencies using nearest neighbor validation results. We were able to discern four uncertainty classes with values of 7%, 18%, 20% and 43%, which we called “ground truth probabilities”. We present the methods to create these models as a specific example of how disaggregation techniques may be used to aid in updating national soil survey inventories.

Domain driven data mining in human resource management: A review of current research

An increasing number of publications concerning data mining in the subject of human resource management (HRM) indicate the presence of a prospering new research field. The current paper reviews this research on HR data mining to systematically uncover recent advancements and suggest areas for future work. Based on the approach of domain driven data mining, an initial framework with significant domain-specific requirements is elaborated. Relevant research contributions are identified and reviewed against the background of this framework. The review reveals that HRM constitutes a noteworthy new domain of data mining research that is dominated by method- and technology-oriented work. However, specific domain requirements, such as evaluating the domain success or complying with legal standards, are frequently not recognized or considered in current research. Therefore, the systematic consideration of domain-specific requirements is demonstrated here to have significant implications for future research on data mining in HRM.

New insights into churn prediction in the telecommunication sector: A profit driven data mining approach

Customer churn prediction models aim to indicate the customers with the highest propensity to attrite, allowing to improve the efficiency of customer retention campaigns and to reduce the costs associated with churn. Although cost reduction is their prime objective, churn prediction models are typically evaluated using statistically based performance measures, resulting in suboptimal model selection. Therefore, in the first part of this paper, a novel, profit centric performance measure is developed, by calculating the maximum profit that can be generated by including the optimal fraction of customers with the highest predicted probabilities to attrite in a retention campaign. The novel measure selects the optimal model and fraction of customers to include, yielding a significant increase in profits compared to statistical measures. In the second part an extensive benchmarking experiment is conducted, evaluating various classification techniques applied on eleven real-life data sets from telecom operators worldwide by using both the profit centric and statistically based performance measures. The experimental results show that a small number of variables suffices to predict churn with high accuracy, and that oversampling generally does not improve the performance significantly. Finally, a large group of classifiers is found to yield comparable performance.

Expression Detection Based on a Novel Emotion Recognition Method

As facial expression is an essential way to convey human's feelings, in this paper, a dynamic selection ensemble learning method is proposed to analyze their emotion automatically. A feature selection algorithm is proposed at first based on rough set and the domain oriented data driven data mining theory, which can get multiple reducts and candidate classifiers. Then the nearest neighborhood of each unseen sample is found in a validation subset and the most accurate classifier is extracted from the candidate classifiers. Finally, the selected classifier is used to recognize unseen samples. Experimental results show that the proposed method is effective and suitable for emotion recognition.

DDDM2007

Real-world data mining generally must consider and involve domain and business oriented factors such as human knowledge, constraints and business expectations. This encourages the development of a domain driven methodology to strengthen data-centered pattern mining. This report presents a review of the ACM SIGKDD Workshop on Domain Driven Data Mining (DDDM2007), held in conjunction with the Thirteenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD07), which was held in San Jose, USA on 12 August, 2007. The aims and objectives of this workshop were to provide a premier forum for sharing innovative findings, knowledge, insights, experiences and lessons in tackling challenges met in domain driven, actionable knowledge discovery in the real world.

Integrative genomics based identification of potential human hepatocarcinogenesis-associated cell cycle regulators: RHAMM as an example

DNA microarray has been widely used to examine gene expression profile of different human tumors. The information generated from microarray analysis usually represents the overall range of cancer-associated abnormality associated with gene regulation. In order to identify key regulatory genes involved in carcinogenesis of human cancer, hypothesis driven data mining of the microarray data plus experimental validation becomes a critical approach in the post-genome era. Here, we present an integrative genomic analysis of published microarray data and homolog gene database. Over 20,000 genes were examined to reveal 16 genes specific to vertebrates, cell cycle G2/M regulated, and overexpressed in human HCC. Using Affymetrix microarray analysis, we found that all 16 genes were up-regulated in human HCC. Among these 16 genes, we experimentally validated the up-regulation of receptor for hyaluronan-mediated motility (RHAMM) in different cell model systems. We first confirmed elevation of RHAMM in the G2/M phase of synchronized HeLa cells. We also found that RHAMM had an elevated level of expression in all the HCC samples we examined and it was induced during the G2/M phase of regenerating mouse hepatocytes after partial hepatectomy. Thus, the expression of RHAMM appears to be tightly regulated during mammalian cell cycle G2/M progression. The ectopic overexpression of RHAMM in 293T cells resulted in the accumulation of cells at G2/M phase. RHAMM-induced mitotic arrest of cells was predominantly in the prophase. Taken together, using an integrated functional genomic approach, we have uncovered a set of genes that may play specific roles in cell cycle progression and in HCC development. To elucidate the function of these genes in cell cycle regulation may shed light on the control mechanism of human HCC in the future.