E.g., 04/03/2020
E.g., 04/03/2020

KantanMT Launches BuildAnalytics™


Tuesday, 17 December, 2013

KantanMT is pleased to announce the launch of its new training data analysis technology, KantanMT BuildAnalytics™. BuildAnalytics measures the quality of all training data uploaded during the engine building process, and generates graphical diagnostic reports for three Quality Evaluation Metrics; BLEU, TER, and F-Measure. This offers greater transparency and control to members when building customised KantanMT engines.

“The KantanMT BuildAnalytics feature is breaking new ground within the Localization industry,” said Eric Chubb, Senior Software Developer, KantanMT.com. “Taking a granular approach to industry standard metrics of quality evaluation (QE), means that our members are offered greater insight and are more transparency when customising MT engines.”

The BuildAnalytics technology takes approx. 1000 segments, randomly selected from the members uploaded training data, and generates segment level quality scores which are expressed as a percentage. It calculates a BLEU, TER and F-Measure score for each individual segment and graphically displays these score in the user dashboard. The BuildAnalytics report can be accessed through the 'My Jobs' list and is available to download as an Excel file.

For more information, please go to www.KantanMT.com, or contact Niamh Lacy.

What is KantanMT™?

KantanMT is a cloud-based implementation of Moses Statistical Machine Translation technology. Leveraging the power and flexibility of the cloud, KantanMT effortlessly scales to generate a high-quality, low-cost Machine Translation platform for Language Service Providers. According to the research firm, Markets and Markets, the natural language processing sector will grow from €2.8 billion in 2013 to €7.4 billion in 2018, representing a compound annual growth rate of 21.1 percent.

What is Xcelerator?

Xcelerator Machine Translations Solution is a new startup that is focusing on making Machine Translation easy to use, inexpensive to access and available to mid-sized LSPs that are eager to integrate MT into their clients localization workflows.