[vc_row][vc_column width=”1/6″][/vc_column][vc_column width=”2/3″][vc_column_text]

eContext White Paper Explores Machine Learning’s Potential through Classified Context

[/vc_column_text][/vc_column][vc_column width=”1/6″][/vc_column][/vc_row][vc_row][vc_column][vc_column_text]

White paper shows how taxonomy can help machine learning deliver more relevant business insights

[/vc_column_text][/vc_column][/vc_row][vc_row css=”.vc_custom_1469563988030{padding-top: 30px !important;}”][vc_column width=”1/6″][/vc_column][vc_column width=”2/3″][vc_column_text]Chicago (June 7, 2016) – eContext (www.econtext.ai), the world’s largest semantic text classification engine, released its latest white paper today, Contextual Machine Learning: It’s Classified, authored by leading text analytics expert Seth Grimes. The whitepaper explores the foundations of machine learning built upon the power of established, proven classification and analysis methods and discusses the advantages of pattern detection and classification at the heart of search, social listening and customer engagement to help deliver the most accurate and relevant insights. “Outcomes are more favorable if models are trained with the help of deep, curated data structures, tuned via active learning, and the results are channeled through a topic hierarchy,” said Stephen Scarr, eContext CEO. “Applying taxonomy to a proven machine learning model provides unbounded data relevancy for the most robust and relevant insights.” With more brands employing chatbots, virtual agents and intelligent assistants to interact with target audiences, the whitepaper explains efficient methodology to train the technology/improve machine learning accuracy with classified context, including:

• Creating a high-relevancy training set by drawing only from sources that provide on-topic inputs, and applying contextual classification to ensure that each input is relevant

• Applying eContext’s deep linguistic library to automate annotation
• Using taxonomy to provide implied annotations (e.g. – labeling a car-model instance with a tag for manufacturer, even when the manufacturer’s name isn’t explicitly present in the training data)
• Checking machine-learning outputs against gold-standard results, which are typically produced by human evaluators
• Keeping models current by using output corrections to adaptively retrain the machine-learning-produced model

To download the white paper, visit: https://www.econtext.ai/MachineLearning. About eContext The world’s largest semantic text classification engine, eContext classifies text in real time to any of its 450,000 topic categories within a hierarchical structure. Powered by a dynamically growing proprietary knowledge base of over 55 million curated rules and billions of documents and online consumer interactions, eContext has invested more than 1 million hours of R&D to date. eContext’s 25 verticals cover all sectors of the consumer digital experience: commerce, chat, social, news, entertainment, and more. eContext’s service accepts inputs from 35 languages, and can be used to topically map path-to-purchase and consumer journey, predict user behavior, classify videos, validate image recognition, and structure training corpuses. eContext also helps clients apply natural-language intelligence to voice-activated assistants and chatbots. Clients include Ask.com, DataSift, Kantar Media, and Publicis Groupe. eContext is owned by metasearch company Info.com. For more information, visit www.econtext.ai.[/vc_column_text][/vc_column][vc_column width=”1/6″][/vc_column][/vc_row][vc_row css=”.vc_custom_1469564147894{padding-top: 30px !important;padding-bottom: 75px !important;}”][vc_column width=”1/6″][/vc_column][vc_column width=”2/3″][vc_column_text]—###— Contact Rey Perez PReturn Inc. 312-226- 4139 [email protected][/vc_column_text][/vc_column][vc_column width=”1/6″][/vc_column][/vc_row]