Predicting Product Recall by Using Machine Learning to Analyze Customer Reviews
Open AccessProduct quality engineers and product managers are routinely seeking ways to improve the identification of products that have the potential to be recalled. Recalled products can cause injury to users, damage to property, loss of revenue, loss of customer loyalty, damage to brand reputation, and product liability suits. Delays in identifying and executing product recalls only serve to exacerbate each of the above outcomes, as the product remains in distribution for a longer period of time.Engineers can use online customer reviews as a source of information for identifying products that have the potential to be recalled. However, large volumes of reviews and products make it difficult to analyze all of them in a timely manner. This research investigates the use of machine learning to automate the analysis of reviews and to improve the timeliness of identifying potential recall products. Similar research has largely concentrated on identifying reviews that contain defect information, or on determining the sentiments expressed in the reviews. This research goes further by relating the information contained in customer reviews to product recalls, and by determining the extent to which earlier potential product recalls could have been identified. Other research into machine learning for product reviews has shown success using multiple Recurrent Neural Networks (RNN) in combination with the Latent Dirichlet Allocation (LDA) topic model. For this praxis, multiple Convolutional Neural Network models are developed to first classify reviews with safety related information, then identify potential recall products based on network training. The praxis is applied more specifically to data from consumer products sold online or in traditional brick-and-mortar store outlets.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
Williams_gwu_0075A_15671.pdf | 2022-03-06 | Open Access |
|