Over 50% Translation Cost Reduction with YarakuZen
IDS Co. Ltd
Industry
Logistics · Food
Use
Big Data Translation
Introduction Purpose
To make product information and big data more than 4 million items which are more multilingual
Goal
Cutting cost of translation
Introduction Effect
Dictionaryization of product information in various languages, translation cost cut by 50%
Reference Plan
First Attempt to unify Food Data in Japan
“I-code” is the first product standardization code in Japan that covers up to “fresh three goods and prepared foods” that were impossible to unify in the food distribution industry, and unified all products handled into data that can be analyzed and utilized. Japan Distribution Science & Information Center (JDIC) and IDs Corporation are the first companies that develop “i-code” and offer the services used for promotional activities. In addition, i-code is also used in POS data of one-third of supermarkets in Japan including AEON Group, and it analyzes not only the B to B marketing tool but also the daily price index measurement led by the Ministry of Internal Affairs and Communications and the University of Tokyo.
Multilingualizing Product Information of 4 million Commodities
“Currently, i-code takes charge of approximately 3,500 stores data operated by around 60 firms in Japan. Since the size of data is so large that the amount of product data has already been 4 million, we can discover a huge opportunity to utilize i-code to satisfy the inbound demand.” Yoshiaki Kawahira, Managing Director of JDIC, has been developing the devices that display product information in their native language when those foreigners scan the product, and the platform that acquires the data on consumer behaviors based on these devices. “In recent years, the number of foreign tourists and residents is increasing in Okinawa. We have sensed the value of utilizing data log of those consumer behaviors for marketing support and have endeavored to create the platforms. The foreigners consumer behavior is different from Japanese, there is a tendency for customers to examine the product data in detail, so we have also developed a device that displays product information.But on the other hand, the enormous Japanese How to multilanguage the product data of the product has become a big issue.”
Up to 50% Improvement of the Efficiency of Big Data Translation
“The first important thing we consider is the need of a tool that reflects on the translation units in machine. In our case, since the Japanese data that need to be multilingualized are too large, the modification of fuzzy translation by human is almost impossible. On the other hand, there are also some articles that are hardly translated by machine. However, YarakuZen solves all these problems. YarakuZen automatically displays the registered translation units in machine translation, which exactly meets what we desire. We are also attracted by the simple operation platform that enables us to easily request human translation from the page of machine translation.”, said by Kawahira. He has introduced YarakuZen into the current translation team and used it mainly for the translation of product data.
As a Translation Tool for the Latest Century
“Machine translation can translate up to 50,000 characters at a time, so it is convenient to translate a huge amount of data at once. Moreover, since whether the translation results match with registered translation units is distinguished by different colors, it is available for us to recognize the translation contents by our senses. In terms of customer support, the person in charge will explain carefully, so we often talk to them when we are in trouble of using it.”, said by Hira Idees. We are using YarakuZen as a “multilingual version dictionary” of huge product data at the business site, and we plan to use it for multilingualization of new services following inbound demand.