Back to the newsroom

Press Release

Korea Deep Learning uses DEEP OCR to read commerce product detail pages and turn them into product databases

Korea Deep LearningPublished

Korea Deep Learning uses DEEP OCR to read commerce product detail pages and turn them into product databases
In September 2021, Korea Deep Learning unveiled a solution that uses DEEP OCR to read the prices, ingredients and product features on commerce product detail pages and build them into a product database.

On September 13, 2021, Korea Deep Learning unveiled a solution that reads commerce product detail pages with AI and collects product data automatically. Using DEEP OCR, its own optical character recognition (OCR) technology, it turns detail page images into text and builds information such as prices, ingredients and product features into a database (DB). The solution saves sellers and platforms the effort of copying product information by hand.

Detail pages that mix text, images and tables

Product detail pages are a typical form of unstructured content, mixing text, images and tables. Decorative lettering, design fonts, certification marks and ingredient tables are combined in a single image, so sellers and platforms have long had people copy product information out one item at a time.

Korea Deep Learning's solution recognizes not only the Korean text, special symbols and tables on detail pages but also special certification marks, decorative lettering inside images, handwriting and a wide range of design fonts, slanted characters and complex layouts. For a product from a food shopping mall, for example, it extracts information such as price, ingredients, product features, how to take it and expiration date as text data. The recognized text goes through a verification process to raise accuracy and reliability, and it can be linked automatically to business systems.

Data for sales strategy and information platforms

The extracted data can be used by sellers to analyze products, plan sales strategies and follow market trends. On platforms that gather and present information about specific product categories, it is used to provide detail pages and ingredient tables as text.

"As e-commerce has developed, consumers can now compare a wide range of products in one place. That makes it essential for sellers to build their selling points and keep track of market trends."

— Kim Ji-hyun, CEO, Korea Deep Learning

Kim Ji-hyun, CEO of Korea Deep Learning, explained that DEEP OCR, based on AI character recognition, will help boost product sales by giving shopping mall sellers the data on their product pages as a DB they can use in their sales activities.

DEEP OCR extends from documents to commerce

DEEP OCR is a real-time character recognition technology that reads the text in documents with the company's own deep learning engine. Korea Deep Learning unveiled it in June 2021, highlighting an accuracy of 99.9%. It is designed to work even on a company's closed internal network without an internet connection, so it can be used by institutions with strict security requirements.

Korea Deep Learning has built systems including an ultra-high-speed logistics OCR system for Korea Post, a biometric immigration system for Gimpo International Airport and a system for digitizing radiation exposure records for the Korea Centers for Disease Control and Prevention. It is a case of recognition technology that began with documents and images extending its reach into commerce data.

About DEEP OCR

Korea Deep Learning started out by reading documents. Built on KDL's own document-specialized VLM, DEEP OCR is Korea's first VLM-based OCR solution, reading not just the characters but the structure and meaning of a document together. It handles unstructured documents in every kind of layout without building separate training data, and extracts text and key-value data from visual elements such as seals, signatures, tables, numbers and symbols. From loan documents in finance to administrative documents in the public sector and handwritten waybills in logistics, documents people used to copy out by hand become business data through DEEP OCR. More information is available on the DEEP OCR page at koreadeep.com.

About Korea Deep Learning

Korea Deep Learning (KDL), founded in 2019, is an enterprise AI agent company. Built on KDL Frontier, its own vision-language model (VLM) specialized for documents, KDL makes AI agents that take on and complete the document work of businesses and public institutions. Its main products are DEEP OCR, DEEP Parser, DEEP Agent and DEEP Ops.

Source