Explorance MLY: Turn Qualitative Feedback into Insights with Machine Learning

Uploaded comments are only retained for 2 weeks in the MLY back-end before deletion. This measure is taken for support and quality assurance purposes only. Comments are stored on Microsoft Azure infrastructure in the U.S.

Your data is your data. Comments uploaded to MLY are not used for any purpose beyond the requested analysis. In addition, customer data is only used in training MLY if the customer has provided written consent.

MLY analyzes qualitative data and summarizes that data into quantitative feedback and insights such as sentiments, alerts, recommendations, and topics. Using Natural Language Processing (NLP), MLY can recognize hidden patterns and correlations in the data, cluster and classify them, and improve as it processes more data over time.

Explorance uses a supervised machine learning approach to train MLY. This methodology ensures MLY is trained only on comments formally approved through an in-house blind annotation process. This process uses three annotators working independently, and a comment is approved only if all three unanimously agree on the interpretation.

MLY’s primary strengths are its specialized categorization and actionable insights. Built to understand the student and employee experience, the analysis produces more targeted, relevant insights with themes and terminology specific to the topic. Additionally, MLY enables decisive action through its Recommendations and Alerts models, providing a starting point for the most critical themes in the data.

MLY can analyze feedback sources such as engagement surveys, course evaluations, performance reviews, experience surveys, peer reviews, program evaluations, social media, review websites, discussion forums, and more.

During an analysis, each comment is given an alert score between 0 and 100. The system compares that score to the alert threshold and displays any scores at or above the threshold as an alert in the results. The default threshold is 50 but can be adjusted higher or lower to match your organization's policies, tolerance, or culture regarding which comments should be reviewed for potential follow-up.

Multiple glossaries can be created to accommodate the use of specialized acronyms or abbreviations in different departments, faculties, locations, etc., and then applied to those comments.