Technology is described herein for generating questions using a neural network. The technology generates the questions in a three-step process. In the first step, the technology selects, using a first neural network, a subset of textual passages from an identified electronic document. In the second step, the technology generates, using a second neural network, one or more candidate answers for each textual passage selected by the first neural network, to produce a plurality of candidate passage-answer pairs. In the third step, the technology selects, using a third neural network, a subset of the plurality of candidate passage-answer pairs. The technology then generates an output result that includes one or more output questions chosen from the candidate passage-answer pairs produced by the third neural network. The use of the first neural network reduces the processing burden placed on the second and third neural networks. It also reduces latency.

Technology is described herein for generating questions using a neural network. The technology generates the questions in a three-step process. In the first step, the technology selects, using a first neural network, a subset of textual passages from an identified electronic document. In the second step, the technology generates, using a second neural network, one or more candidate answers for each textual passage selected by the first neural network, to produce a plurality of candidate passage-answer pairs. In the third step, the technology selects, using a third neural network, a subset of the plurality of candidate passage-answer pairs. The technology then generates an output result that includes one or more output questions chosen from the candidate passage-answer pairs produced by the third neural network. The use of the first neural network reduces the processing burden placed on the second and third neural networks. It also reduces latency.

Methods, apparatuses, and storage media storing instructions for classifying text documents are provided. A plurality of text documents is obtained. The plurality of text documents is classified into one or more document categories based on a plurality of classification rules. Each of the one or more document categories include one or more first text documents of the plurality of text documents. A second text document of the plurality of text documents is classified based on the plurality of classification rules as belonging to none of the one or more document categories. One or more document fingerprints are generated for respective first text documents in the one or more document categories. The second text document is classified into one of the one or more document categories based on the one or more document fingerprints.

Methods, apparatuses, and storage media storing instructions for classifying text documents are provided. A plurality of text documents is obtained. The plurality of text documents is classified into one or more document categories based on a plurality of classification rules. Each of the one or more document categories include one or more first text documents of the plurality of text documents. A second text document of the plurality of text documents is classified based on the plurality of classification rules as belonging to none of the one or more document categories. One or more document fingerprints are generated for respective first text documents in the one or more document categories. The second text document is classified into one of the one or more document categories based on the one or more document fingerprints.

Disclosed are a character string transmission method and device, a computer, and a readable storage medium. The method includes the following steps: obtaining a target character string, and adding an escape character before each special character in the target character string, the special character being a character that is incapable of being transmitted accurately to a target script; converting the special character into a transcoded character in an American Standard Code for Information Interchange (ASCII) code form to obtain a transcoded character string; transmitting the transcoded character string to the target script by means of a shell; and calling the target script to decode the transcoded character string to obtain the target character string. Compared with existing complex escape, the method is easier to implement, and the special character may be effectively prevented from being specially processed by the shell.

Disclosed are a character string transmission method and device, a computer, and a readable storage medium. The method includes the following steps: obtaining a target character string, and adding an escape character before each special character in the target character string, the special character being a character that is incapable of being transmitted accurately to a target script; converting the special character into a transcoded character in an American Standard Code for Information Interchange (ASCII) code form to obtain a transcoded character string; transmitting the transcoded character string to the target script by means of a shell; and calling the target script to decode the transcoded character string to obtain the target character string. Compared with existing complex escape, the method is easier to implement, and the special character may be effectively prevented from being specially processed by the shell.

Embodiments of the present disclosure provide a method, an electronic device, and a computer program product for training a model. The method may include determining image features, audio features, and text features of a reference object based on reference image information, reference audio information, and reference text information associated with the reference object, respectively. The method may also include constructing a feature tensor from the image features, the audio features, and the text features. In addition, the method may further include decomposing the feature tensor into a first feature vector, a second feature vector, and a third feature vector corresponding to the image features, the audio features, and the text features, respectively, to determine a loss function value of the model. The method may also include updating parameters of the model based on the loss function value.

Embodiments of the present disclosure provide a method, an electronic device, and a computer program product for training a model. The method may include determining image features, audio features, and text features of a reference object based on reference image information, reference audio information, and reference text information associated with the reference object, respectively. The method may also include constructing a feature tensor from the image features, the audio features, and the text features. In addition, the method may further include decomposing the feature tensor into a first feature vector, a second feature vector, and a third feature vector corresponding to the image features, the audio features, and the text features, respectively, to determine a loss function value of the model. The method may also include updating parameters of the model based on the loss function value.

Methods and systems for unsupervised training for a neural multilingual sequence-to-sequence (seq2seq) model. Denoising adapters for each of one or more languages is inserted into an encoder and/or a decoder of the seq2seq model. Parameters of the one or more denoising adapters are trained on a language-specific denoising task using monolingual text for each of the one or more languages. Cross-attention weights of the seq2seq model with the trained denoising adapter layers are fine-tuned on a translation task in at least one of the one or more languages with parallel data.

Methods and systems for unsupervised training for a neural multilingual sequence-to-sequence (seq2seq) model. Denoising adapters for each of one or more languages is inserted into an encoder and/or a decoder of the seq2seq model. Parameters of the one or more denoising adapters are trained on a language-specific denoising task using monolingual text for each of the one or more languages. Cross-attention weights of the seq2seq model with the trained denoising adapter layers are fine-tuned on a translation task in at least one of the one or more languages with parallel data.