Aspectiva extracts meaningful insights out of reviews using advanced NLP approaches. We focus on the sentiment in each review, as well as different product aspects, and generate useful observations based on them. As reviews are written in free text, we first need to transform them into something the computer can understand. We inspect word-level and sentence-level characteristics, deducing valuable morphological and syntactic information. We then apply our profound semantic computational linguistic knowledge to automatically comprehend the review, saving retailers and customers hours of research work.
We combine our highly accurate extraction capabilities with cutting-edge, transformer-based deep learning models to further enrich and expand the information we extract from each review. As we’re dealing with large amounts of data and users, we carefully design and implement different pipelines to support the high traffic going through Walmart.com.
The various insights we extract from reviews, alongside other data sources, manifest in different features within Walmart.com. Shoppers can focus on the aspects they care about most, making smart, informed decisions. And retailers can get valuable, concentrated insights on each and every one of their products.