The following text can be used as a partner description of the company for a grant or R&D proposal.
Pangeanic is a is a translation technology developer and language service provider with a strong focus on machine translation applications for processing multilingual big data. The company was mentioned by the Euromatrixplus.net FP7 project as the “first company in the world to apply Moses successfully” after making use for the first time of that EU programme’s results in a commercial setting. Since then the company has developed a number of hybrid machine translation commercial tools and an automated translation platform, which self-learning and auto-retraining capabilities. The automated translation platform will feed directly into the objective of WASABI. Pangeanic’s machine translation platform contains powerful format filters and re-training features. This ensures that several input formats can be used in order to obtain final machine translation output. Powerful APIs can be used to link web data to machine translation and human post-editing services.
Pangeanic currently participates in a Marie Curie action: EXPERT led by the Univesity of Wolverhampton. EXPERT (http://expert-itn.eu/). This program aims to promote the research, development and use of hybrid language translation technologies. The overall objective of EXPERT is to provide innovative research and training in the field of Translation memory and Machine Translation Technologies. Pangeanic has participated in 3 regional research programs with the local Computer Science Institute (ITI) since 2009 to develop and improve its machine translation technology, which has received FEDER funds, as well as European funds for the taking on of a researcher (2010-2014) in machine translation. The company has made public its research via publications.
The creation of an automated re-training platform using Moses was a world-first and it was published by AMTA (Association for Machine Translation in the Americas) https://www.pangeanic.com/news/machine-translation-technology-implementations-amta/ and http://www.euromatrixplus.net/moses-decoder/ Later on, the company has undertaken a program with Toshiba to research improvement on hybridisation on Japanese, which would also benefit non-related language pairs. This collaboration lead to two further publications at AAMT (Asian Association for Machine Translation) http://blog.pangeanic.com/2012/01/20/toshiba-and-pangeanic-steps-to-machine-translation-hybridation-article-in-aamt/ (downloadable) or http://www.pangeamt.com/demo/wp-content/uploads/2012/01/Pangeanic_Toshiba_Hybridation_AAMT.pdf second article (downloadable): https://www.pangeanic.com/news/diy-machine-translation-aamt/ The company is involved in the application of machine translation output applied to the translation of big data in private settings for social media analysis/sentiment. In this respect, future work will concentrate on the application of machine translation for fast multilingual web generation, translation of non-standard sentences coming from social media and, as an area of key business development, multilingual data retrieval and multilingual search with no-geolocation in order to provide the most relevant search results, irrespective of language.
Position on the value chain
Multilingual support, machine translation
Main contribution for the project
Pangeanic will utilise its technology to provide multilingual support to the project. It will develop, test and create machine translation engines capable of transferring big data and inputs from several languages into a target language. This will provide the basis for multilingual sentiment analysis and multilingual communication.
Key: machine translation (automated translation of inputs and outputs by users), translation services, translation of massive amounts of data (Big Data).
Companies such as Sony, Sybase, Honda, Panasonic, Ciriec already benefit from our machine translation and translation API solutions. As the brand is already established and generating revenue, Pangeanic intends to expand its growth in the multilingual search area, focusing growth towards its traditional markets in Japan, China and Europe, with the US in a second phase, particularly in the speech-to-text area. Due to the potential of our technology, enhancing multilingual communication and fast translation of big data Pangeanic targets companies that own large amounts of multilingual content, sentiment analysis companies that need to process multilingual information (what do Italians think about Merkel’s reelection in social media) and, very importantly, search engines to retreive content which cannot be found as queries are monolingual in all search engines. Typical customers are language service providers for translation engines, sentiment analysis companies, and search companies. All these companies can strongly benefit from a multilingual approach to handling big data, improving the maximizing the data and information they handle.