Walmart Inc.
Systems and methods for a customer feedback classification system

Abstract:

Systems, methods, and machine readable media are provided for classifying customer feedback. In exemplary embodiments, text is captured from at least one source relating to at least one product. The text is scanned and a score is produced for sentiment for the at least one product. The text is filtered into parts of speech and key words to produce filtered text. The filtered text is transformed into a term-document matrix. A risk score is calculated and prioritized based on the term-document matrix and the sentiment score. The product and the associated risk score are reported to a subject matter expert (SME), where a determination is made whether the product is reportable or non-reportable.