Introducing Neoscore on Neofiliac
Neofiliac is introducing Neoscore for product pages from today. Based on deep learning, Neoscore produces a rating for each product by analyzing reputable third-party reviews and other public sources of information. This score reflects the overall reviewer perception of each product, making it easy to get a general idea without delving into often dozens of review articles.
Neoscore appears in several places, including Product Comparison
Why Neoscore?
Today, there is no shortage of review articles on popular products. While each reviewer has a different approach with diverse methodologies, these reviews on the same products show clear trends: Some are generally loved; some are generally hated; some receive a mix of both. At each price point, with similar specifications, different models from different manufacturers can fall anywhere along this spectrum. Having a way to measure the overall perception of the product while being relatively safe from "flukes" or outliers is advantageous for consumers.
Under the bonnet, Neoscore uses Natural Language Processing (NLP) to understand what reviewers mean when they point out the pros and cons of each product. This allows it to form an opinion of the product in the form of a score.
At the same time, not all products are popular with reviewers. It can be frustrating trying to find objective third-party reviews on such products. In a future update, Neoscore will use more data sources to predict a score for products that have yet to be reviewed (if they ever will be).
Neoscore joins the product card as a new component at its bottom-right corner
Next Step: Reviewer AI
In the coming weeks, we will also introduce a reviewer AI that produces abstracts (pros and cons) for products to complement Neoscore. It will cover a different aspect of product comparison: details that reviewers have noticed but cannot be reflected in a single score. Together, they serve to help consumers get the gist of dozens of review articles in under a minute.
Article Gallery
