Data Mining Workbench Research Articles

Automated Dataset Construction from Web Resources with Tool Kayur

Many text mining tools cannot be applied directly to documents available on web pages. There are tools for fetching and preprocessing of textual data, but combining them with the data processing tool into one working tool chain can be time consuming. The preprocessing task is even more labor-intensive if documents are located on multiple remote sources with different storage formats. In this paper, we propose the simplification of data preparation process for cases when data come from wide range of web resources. We developed an open-source tool, called Kayur, that greatly minimizes time and effort required for routine data preprocessing steps, allowing to quickly proceed to the main task of data analysis. The datasets generated by the tool are ready to be loaded into a data mining workbench, such as WEKA or Carrot2, to perform classification, feature prediction, and other data mining tasks.

Detecting Anomalies in Students‟ Results Using Decision Trees

Examinations are one of the most important activities that take place in institutions of learning. In many Nigerian universities, series of meetings are held to manually examine and approve computed student examination results. During such meetings, students results are scrutinized. Reasonable explanations must be provided for any anomaly that is discovered in a result before the result is approved. This result approval process is prone to some challenges such as fatigue arising from the long duration of the meetings and wastage of man- hours that could have been used for other productive tasks. The aim of this work is to build decision tree models for automatically detecting anomalies in students examination results. The Waikato Environment for Knowledge Analysis (WEKA) data mining workbench was used to build decision tree models, which generated interesting rules for each anomaly. Results of the study yielded high performances when evaluated using accuracy, sensitivity and specificity. Moreover, a Windows-based anomaly detection tool was built which incorporated the decision tree rules. Index Terms—Decision trees, examination results, anomaly detection, educational data mining, result anomaly.

Decision Model for Acute Appendicitis Treatment With Decision Tree Technology—A Modification of the Alvarado Scoring System

How to decide the proper time to do laparotomies for acute appendicitis patients is sometimes very difficult, especially in areas with no imaging diagnostic tools. The Alvarado scoring system (ASS) is a convenient and inexpensive decision making tool; however, its accuracy needs to be improved. The decision tree is the most frequently used data mining technology for diagnostic model building. This study used a decision tree to modify the ASS and to prioritize the variables. We collected 532 patients who underwent appendectomy. Patients who had undergone incidental appendectomy were excluded from the study. The decision tree algorithm was constructed with the data mining workbench Clementine version 8.1. It is a top-down algorithm designed to generate a decision tree model with entropy. The algorithm chooses the best decision node with which to separate different classes from empirical data. The Wilcoxon signed rank test, Student t test and chi(2) test were used for statistical analysis. Among the 532 patients recruited into the study, 420 had acute appendicitis and 112 had normal appendix. Women with acute appendicitis were older than their male counterparts (p < 0.001). All patients had right lower quadrant tenderness. The new model was constructed with decision tree technology, and the accuracy of the diagnostic rate was better than that of ASS (p < 0.001). The sensitivity and specificity of the new model were 0.945 and 0.805, respectively. The new model is more convenient and accurate than ASS. Right lower quadrant tenderness is an inclusion criterion for acute appendicitis diagnosis. Migrating pain and neutrophil count > 75% were significant factors for acute appendicitis diagnosis if ASS score < 6. Although the criteria of nausea/vomiting and white blood cell count > 10,000/dL were significantly different between acute appendicitis and normal appendix, there was no significant contribution of entropy change below the "neutrophil count > 75%" nodes in the model. So they were erased from the decision tree model. Further studies need to be conducted to investigate why older women are at higher risk for acute appendicitis.

Predicting Crime Scene Attendance

This paper ascertains the feasibility of using a data mining supervised learning algorithm to predict which crime scenes potentially offer the best opportunity of recovering forensic samples such as finger prints or DNA. UK police forces have a finite number of Crime Scene Investigators and an ever-increasing demand to attend crime scenes. Most forces have a documented attendance criterion but crimes are not evenly distributed throughout the working day. This, at times, results in more scenes to examine than there are investigators, thereby posing the question, ‘Which scenes should be attended first in order to retrieve forensics?’. The Insightful Miner data mining workbench tool has been used to retrieve, combine, clean, manipulate and model the data. The results from Northamptonshire Police have demonstrated that the modelling process is able to predict crime scene attendance to an accuracy of 68 per cent which is significantly better than human experts. The methodology was used in a second police force (Gwent) which records data into different computer systems and the results were comparable to those achieved in Northamptonshire Police. This demonstrates that the methodology is portable between forces.