The automatic summary is addressed to the sales teams or customer service in organizations in order to digest quickly the emails and other information, allowing them to dedicate more time to their customers.

This year it is hoped that people spend more of half a day reading emails, articles or publications in social networks, and this will go to more.

to help solve the time that we must devote to the huge amount of information with which we move, the researchers Salesforce have developed an algorithm that uses the machine learning to produce summaries “consistent and accurate”, according to the MIT Technology Review.

The automatic summary will be a technology particularly useful for Salesforce, which produces a variety of products oriented to customer service. The company notes that the summaries resulting could be used by the equipment sales or service to the client to digest information quickly, allowing them to devote more time to their customers.

We used machine learning to find ways of summarizing longer blocks of text, that I could eventually incorporate into their products.

The company has announced that it has carried out two major advances in the natural language processing introducing a new “model of generation of words of context-sensitive” and a “new way to create models summary”. Together, the two advances allow researchers to automatically create summaries of longer texts, accurate and legible. is The company acquired equipment in the company of deep learning (deep learning) MetalMind last year, that is behind the research.

The researchers explain that the automatic summary of the text works in two ways: extraction or abstraction. With the removal, the computer may draw from a pre-existing text. The abstraction allows the computer to enter new words, but the system has to understand the original article enough as to be able to introduce the right words.