[ad_1]
Samsung Research, Samsung Electronics' advanced R & D center of SET business (finished products), is ranked first of two of the world's largest international reading comprehension competitions. artificial intelligence (AI).
in the MAchine Reading COmprehension (MS MARCO) held by Microsoft (MS), as well as the best performance in TriviaQA hosted by the University of Washington, proving the excellence of its AI algorithm.
With intense competition in the development of AI technologies on a global scale, reading comprehension competitions like MS MARCO are in full swing around the world. MS MARCO and TriviaQA are part of the reading comprehension competitions actively researched and used with SQuAD from Stanford University and DeepMind's NarrativeQA. World-renowned universities and global AI firms, including Samsung, are competing in these challenges.
Reading comprehension is an AI algorithm that badyzes data and finds an optimal response to a query. For MS MARCO and TriviaQA, AI algorithms are tested in their natural language processing capabilities in human questions and answers and also provide written text in various types of documents such as newspaper articles and blog articles.
are presented for some query to let an AI algorithm create an optimal response. Queries are randomly selected from a million queries from Bing users (MS search engine). Responses are evaluated statistically by estimating how close they are to human responses. This is a test designed to apply an AI algorithm to solve real world problems.
Samsung Research participated in competitions with ConZNet, an AI algorithm developed by the AI Center of the company. ConZNet presents skillful abilities by adopting the reinforcement learning technique, which advances artificial intelligence by giving reasonable feedback on the results, similar to a stick and carrot strategy in a process. learning
. , contests are widespread in the areas of computer vision (technologies for badyzing characters and images) and visual questions and answers to solve problems using recognized character images as well as reading comprehension. The Beijing section of Samsung Research won in March the International Conference on Analysis and Recognition of Documents (ICDAR) organized by the International Pattern Recognition Association (IAPR), which allowed it to become a leading group. ICDAR is the most influential competition in Optical Character Recognition (OCR) technologies.
"We are developing an AI algorithm to respond to user queries in a simpler and more convenient way," said Jihie Kim. , Head of the language comprehension laboratory at Samsung Research. "An active discussion is underway at Samsung to adopt the ConZNet AI algorithm for products, services, customer response and technology development."
Click on Deccan Chronicle Technology and Science for the latest news and reviews. on Facebook, Twitter .
[ad_2]
Source link
