October 2017 Dinner Meeting Announcement
October Dinner Meeting – Quality Tools used in Text Mining and Analytics
Information on the Tuesday, October 10, 2017 Dinner meeting has been posted:
Venue: Conference Center at the Maritime Institute, MD
Program: A popular rule of thumb suggests that 80% of the useful data from Voice-of-the-Customer (VOTC) surveys comes from unstructured, free-text, feedback comments. Unstructured text is largely ignored because the data are too cumbersome to analyze. Many of the statistical quality tools for analyzing unstructured text are not as well known to quality professionals, even though they have been widely-used since the early 2000s. Effective applications of these text mining tools include: uncovering root causes of error from comments in engineering hardware and software repair logs, warranty reports, and patient diagnoses and treatment outcomes. This program and tutorial will review and apply the quality tools of text mining and analytics on free-text comments made by attendees of the 2016 ASQ-Baltimore Section meetings. Applying these tools turns unstructured text data into quantitative measures that identify meeting aspects more analytically and reveal new insights for helping section leadership plan better section meetings. .
Presenter: Mel Alexander is an Operations Research Analyst with the Social Security Administration. He uses Statistical Analytics to ensure proper payments go to eligible beneficiaries. Mel also is a Statistical Consultant in the Departments of Diagnostic Radiology and Neurosurgery at the University of Maryland Medical Center where he builds statistical models that identify key factors to increase patient’s chances of surviving traumatic head, spinal cord, and abdominal injuries. He is a past chair of ASQ’s Baltimore Section and Healthcare Division and is a Standing Review Board member for ASQ Quality Press. He holds a master’s degree in Biostatistics from the University of North Carolina. He is an ASQ Fellow and Certified Quality Engineer.
Overview: Review Basic Text Mining Quality Tools and Examples
• Surveys: Text Comments made by respondents
• Word Clouds: Pareto analysis alternative displaying the most common words in free text
• Document-Term-Matrix (DTM): L-shaped Matrix that relates row respondents (documents) to column words (terms) from free-text comments. Cell items (e.g., frequency of terms used) show the strength of relationship between rows and columns.
• Term-Document-Matrix (TDM): transposed matrix of DTM
• Singular Value Decomposition (SVD): matrix algebraic operation that splits DTMs/TDMs into separate matrices expressing numerical relationships between terms and documents that account for most of the information found in unstructured text
• Corpus: Collection of comments or documents found in a Voice-of-the-Customer (VOTC) table
• Hierarchical Clustering: Quantitative form of Affinity Diagram, clustering word groups that convey similar topics or themes.
Click on Events » Dinner Meetings on the website so you can get details on the dinner.
Click on Events » Event Reservations on the website in order to reserve your spot at the dinner.
—
Look forward to seeing you there for networking and a great meeting topic at our new Meeting Venue.
Program Committee