- Number
- 10810382
- Published
- 2020-10-20
- Filed
- 2018-10-09
- Assignee
- Disney Enterprises, Inc.
- Inventors
- Horn; David
- CPC
- G06N3/0464; G06F40/44; G06F40/58; G06N3/08; G06F40/284; G06N3/045; G06F40/216
- Verdict
- Set aside vocabulary/tone conversion NLP tool, business/localization
- Source
- Google Patents · FreePatentsOnline
Abstract
There is provided a content translation system includes a computing platform having a hardware processor and a system memory storing a language conversion software code including a vocabulary conversion convolutional neural network (CNN). The hardware processor is configured to execute the language conversion software code to obtain a content including a language-based content expressed in a first vocabulary. The hardware processor also executes the language conversion software code to convert a wording of the language-based content from the first vocabulary to a second vocabulary using the vocabulary conversion CNN, where the first vocabulary and the second vocabulary are in the same language. The hardware processor further executes the language conversion software code to output a translated content corresponding to the content for rendering on a display, the translated content including the language-based content expressed in the second vocabulary.
Background
BACKGROUND(1) Teens and children may often be curious about subjects that also interest adults, such as science, entertainment, sports, and news about events in the adult world. Although content aggregation services offering newsfeeds including topical information content or content known to be of specific interest to a subscriber exist, those services typically provide content that is developed for adult audiences. As a result, teens and children seeking to learn about or better understand the world around them may find the content provided by conventional content aggregation services to be at too high a reading level to be enjoyable, or to include language that is unsuitable for them. Moreover, much of the news content and commentary generated for adults is uninteresting to a young audience in tone, and may undesirably discourage further inquiry by a young audience.SUMMARY(2) There are provided systems and methods for performing automated conversion of vocabulary and narrative tone, substantially as shown in and/or described in connection with at least one of the figures, and as set forth more completely in the claims.
Claims
1. A content translation system comprising: a computing platform including a hardware processor and a system memory; a language conversion software code stored in the system memory, the language conversion software code including a vocabulary conversion convolutional neural network (CNN); the hardware processor configured to execute the language conversion software code to: obtain a content including a language-based content expressed in a first vocabulary; convert, using the vocabulary conversion CNN, a wording of the language-based content from the first vocabulary to a second vocabulary, wherein the first vocabulary and the second vocabulary are in a same language, wherein the second vocabulary is one of a teen vocabulary or a children vocabulary, and wherein to convert the wording, the vocabulary conversion CNN is configured to: generate a vector corresponding to one of a word or a phrase in the first vocabulary; map the vector to a projection in a vector space corresponding to the second vocabulary; and convert the one of the word or the phrase in the first vocabulary to one of a word or a phrase in the second vocabulary based on the projection; output a translated content corresponding to the converted wording for rendering on a display, the translated content including the language-based content expressed in the second vocabulary.
7. A method for use by a content translation system including a computing platform having a hardware processor and a system memory storing a language conversion software code including a vocabulary conversion convolutional neural network (CNN), the method comprising: obtaining, using the hardware processor, a content including a language-based content expressed in a first vocabulary; converting, using the hardware processor and the vocabulary conversion CNN, a wording of the language-based content from the first vocabulary to a second vocabulary, wherein the first vocabulary and the second vocabulary are in a same language, wherein the second vocabulary is one of a teen vocabulary or a children vocabulary, and wherein the converting includes: generating a vector corresponding to one of a word or a phrase in the first vocabulary; mapping the vector to a projection in a vector space corresponding to the second vocabulary; and converting the one of the word or the phrase in the first vocabulary to one of a word or a phrase in the second vocabulary based on the projection; outputting, using the hardware processor, a translated content corresponding to the converted wording for rendering on a display, the translated content including the language-based content expressed in the second vocabulary.
13. A computer-readable non-transitory medium having stored thereon a language conversion software code including a vocabulary conversion convolutional neural network (CNN) and instructions, which when executed by a hardware processor, instantiate a method comprising: obtaining a content including a language-based content expressed in a first vocabulary; converting, using the vocabulary conversion CNN, a wording of the language-based content from the first vocabulary to a second vocabulary, wherein the second vocabulary is one of a teen vocabulary or a children vocabulary, and wherein the first vocabulary and the second vocabulary are in a same language, wherein the converting includes: generating a vector corresponding to one of a word or a phrase in the first vocabulary; mapping the vector to a projection in a vector space corresponding to the second vocabulary; and converting the one of the word or the phrase in the first vocabulary to one of a word or a phrase in the second vocabulary based on the projection; outputting a translated content corresponding to the converted wording for rendering on a display, the translated content including the language-based content expressed in the second vocabulary.