Efficient and scalable development of multilingual supervised machine learning tools using machine translation and multilingual embeddings

A system comprising:

• one or more processors;
• and
• memory storing thereon instructions that, as a result of being executed by the one or more processors, cause the system to perform operations comprising:
• receiving a source document that includes text segments associated with a source language, wherein one or more of the text segments are associated with a target label;
• translating the text of the source document to generate a set of translated documents, wherein one or more translated documents of the set include text associated with a target language;
• generating a set of labeled multilingual documents, wherein generating the set of labeled multilingual documents includes mapping the target label of the source document to corresponding text segments of the set of translated documents;
• encoding the text of the set of labeled multilingual documents into a plurality of multilingual embeddings;
• and
• training a machine-learning model using the plurality of multilingual embeddings, wherein the machine-learning model is trained to predict whether a particular document includes narrative attacks against one or more entities, and wherein training the machine-learning model includes:
• applying the machine-learning model to the plurality of multilingual embeddings to generate an initial output;
• comparing the initial output with the target label, wherein the target label identifies a classification indicating whether the text of the source language includes the narrative attacks;
• and
• adjusting parameters of the machine-learning model based on the comparison such that the trained machine-learning model generates a target output that corresponds to the target label.

A non-transitory, computer-readable storage medium storing thereon executable instructions that, as a result of being executed by one or more processors of a computer system, cause the computer system to perform operations comprising:

• receiving a source document that includes text segments associated with a source language, wherein one or more of the text segments are associated with a target label;
• translating the text of the source document to generate a set of translated documents, wherein one or more translated documents of the set include text associated with a target language;
• generating a set of labeled multilingual documents, wherein generating the set of labeled multilingual documents includes mapping the target label of the source document to corresponding text segments of the set of translated documents;
• encoding the text of the set of labeled multilingual documents into a plurality of multilingual embeddings;
• and
• training a machine-learning model using the plurality of multilingual embeddings, wherein the machine-learning model is trained to predict whether a particular document includes narrative attacks against one or more entities, and wherein training the machine-learning model includes:
• applying the machine-learning model to the plurality of multilingual embeddings to generate an initial output;
• comparing the initial output with the target label, wherein the target label identifies a classification indicating whether the text of the source language includes the narrative attacks;
• and
• adjusting parameters of the machine-learning model based on the comparison such that the trained machine-learning model generates a target output that corresponds to the target label.