Abstract: The terminal sends user behavior information. The terminal receives a message pushed by a server. The terminal displays, on a first interface, content of the target article information and recommendation information corresponding to the first data. The terminal sends a request of an identifier of the message to the server in response to a first operation on the first interface. The server obtains, based on the request, content of a target article and second data related to the first data. The terminal receives the content of the target article and the second data that are sent by the server. The terminal displays, on a second interface, the content of the target article and the second data that is displayed by using a display template corresponding to the second data.

Abstract: A method for evaluating the mechanical state of a high-voltage shunt reactor based on vibration characteristics is disclosed, relating to the technical field of electrical equipment fault diagnosis. The method includes: based on historical state data and real-time vibration and noise signal data of the high-voltage shunt reactor and through an LSTM neural network time series prediction method, comparing deviation between predicted characteristic value and actual characteristic value, and determining whether the high-voltage shunt reactor has mechanical defects or failures. By using the historical state data and the real-time vibration and noise signal data of the high-voltage shunt reactor, an LSTM neural network time series prediction method, as well as comparison of the deviation between the predicted characteristic value and the actual characteristic value, etc., the evaluation of the mechanical state of the high-voltage shunt reactor is realized.

Abstract: An abstract generation method and apparatus are provided, and relate to the field of artificial intelligence. The method includes: obtaining a content object; obtaining N thumbnails based on N pictures; generating M candidate abstracts based on text information and the N thumbnails; obtaining preference information of a user; selecting one of the M candidate abstracts as an abstract of the content object based on the preference information; and sending the abstract to a terminal device. In the method, a possibility that the user taps the abstract can be increased, and a content object delivery effect can be improved.

Abstract: A network verification system obtains configuration data of a plurality of network devices, where a data model of the configuration data is described by using a general data modeling language independent of the network devices; and the network verification system verifies data links between the plurality of network devices based on the configuration data of the plurality of network devices and a topology structure between the plurality of network devices. The network verification system verifies the data links between the plurality of network devices based on the topology structure between the plurality of network devices and the configuration data described by using the general data modeling language independent of the network devices. This helps improve scalability of the network verification system and avoids relatively poor scalability of network simulation software that occurs when conventional network simulation software provides a template for configuration data of each type of network device.

Abstract: Isolated polypeptides having a heavy chain variable region and/or light chain variable region that specifically binds to CD46 protein as well as antibodies and antibody fragments containing the heavy chain variable region and/or the light chain variable region that bind to CD46 protein. Immunoconjugages, pharmaceutical compositions and kits comprising the polypeptide and antibodies and antibody fragments containing the polypeptide are also provided.

Abstract: Computer implemented method and system for automatic speech recognition. A first speech sequence is processed, using a time reduction operation of an encoder NN, into a second speech sequence comprising a second set of speech frame feature vectors that each concatenate information from a respective plurality of speech frame feature vectors included in the first set and includes fewer speech frame feature vectors than the first speech sequence. The second speech sequence is transformed, using a self-attention operation of the encoder NN, into a third speech sequence comprising a third set of speech frame feature vectors. The third speech sequence is processed using a probability operation of the encoder NN, to predict a sequence of first labels corresponding to the third set of speech frame feature vectors, and using a decoder NN to predict a sequence of second labels corresponding to the third set of speech frame feature vectors.

Abstract: Isolated polypeptides having a heavy chain variable region and/or light chain variable region that specifically binds to Nectin-4 protein as well as antibodies and antibody fragments containing the heavy chain variable region and/or the light chain variable region that bind to Nectin-4 protein. Pharmaceutical compositions and kits comprising the polypeptide and antibodies and antibody fragments containing the polypeptide are also provided.

Abstract: A method includes obtaining n pairs of translation sentences of a source language and a target language, where each of the n pairs of translation sentences includes a source language sentence and a target language sentence that are translations of each other, extracting a source language segment from each source language sentence in the n pairs of translation sentences using an extraction rule of the source language, extracting a target language segment from each target language sentence in the n pairs of translation sentences, and generating an extraction rule of the target language based on n target language segments extracted from n target language sentences.

Abstract: The present disclosure describes methods and systems for generating semantic predictions from an input speech signal representing a speaker's speech, and maps the semantic predictions to a command action that represents the speaker's intent. A streamable multimodal language understanding (MLU) system includes a machine learning-based model, such as a RNN model that is trained to convert speech chunks and corresponding text predictions of the input speech signal into semantic predictions that represent a speaker's intent. A semantic prediction is generated and updated, over a series of time steps. In each time step, a new speech chunk and corresponding text prediction of the input speech signal are obtained, encoded and fused to generate an audio-textual representation. A semantic prediction is generated by a sequence classifier by processing the audio-textual representation and the semantic prediction is updated as new speech chunks and corresponding text predictions are obtained.

Abstract: Methods and systems are described for performing video retrieval together with video grounding. A word-based query for a video is and encoded into a query representation using a trained query encoder. One or more similar video representations are identified, from a plurality of video representations that are similar to the query representation. Each similar video representation represents a respective relevant video. A grounding is generated for each relevant video by forward propagating each respective similar video representation together with the query representation through a trained grounding module. The relevant videos or identifiers of the relevant videos are outputted together with the grounding generated for each relevant video.

Abstract: Methods and systems are described for performing video retrieval together with video grounding. A word-based query for a video is and encoded into a query representation using a trained query encoder. One or more similar video representations are identified, from a plurality of video representations that are similar to the query representation. Each similar video representation represents a respective relevant video. A grounding is generated for each relevant video by forward propagating each respective similar video representation together with the query representation through a trained grounding module. The relevant videos or identifiers of the relevant videos are outputted together with the grounding generated for each relevant video.

Abstract: An abstract generation method and apparatus are provided, and relate to the field of artificial intelligence. The method includes: obtaining a content object; obtaining N thumbnails based on N pictures; generating M candidate abstracts based on text information and the N thumbnails; obtaining preference information of a user; selecting one of the M candidate abstracts as an abstract of the content object based on the preference information; and sending the abstract to a terminal device. In the method, a possibility that the user taps the abstract can be increased, and a content object delivery effect can be improved.

Abstract: Computer implemented method and system for automatic speech recognition. A first speech sequence is processed, using a time reduction operation of an encoder NN, into a second speech sequence that comprises a second set of speech frame feature vectors that each concatenate information from a respective plurality of speech frame feature vectors included in the first set, wherein the second speech sequence includes fewer speech frame feature vectors than the first speech sequence. The second speech sequence is transformed, using a self-attention operation of the encoder NN, into a third speech sequence that comprises a third set of speech frame feature vectors. The third speech sequence is processed, using a probability operation of the encoder NN, to predict a sequence of first labels corresponding to the third set of speech frame feature vectors. The third speech sequence is also processed using a decoder NN to predict a sequence of second labels corresponding to the third set of speech frame feature vectors.

Abstract: A method for evaluating the mechanical state of a high-voltage shunt reactor based on vibration characteristics is disclosed, relating to the technical field of electrical equipment fault diagnosis. The method includes: based on historical state data and real-time vibration and noise signal data of the high-voltage shunt reactor and through an LSTM neural network time series prediction method, comparing deviation between predicted characteristic value and actual characteristic value, and determining whether the high-voltage shunt reactor has mechanical defects or failures. By using the historical state data and the real-time vibration and noise signal data of the high-voltage shunt reactor, an LSTM neural network time series prediction method, as well as comparison of the deviation between the predicted characteristic value and the actual characteristic value, etc., the evaluation of the mechanical state of the high-voltage shunt reactor is realized.

Abstract: A network verification system obtains configuration data of a plurality of network devices, where a data model of the configuration data is described by using a general data modeling language independent of the network devices; and the network verification system verifies data links between the plurality of network devices based on the configuration data of the plurality of network devices and a topology structure between the plurality of network devices. The network verification system verifies the data links between the plurality of network devices based on the topology structure between the plurality of network devices and the configuration data described by using the general data modeling language independent of the network devices. This helps improve scalability of the network verification system and avoids relatively poor scalability of network simulation software that occurs when conventional network simulation software provides a template for configuration data of each type of network device.

Abstract: The method includes: obtaining n pairs of translation sentences of a source language and a target language, where each of the n pairs of translation sentences includes a source language sentence and a target language sentence that are translations of each other; extracting a source language segment from each source language sentence in the n pairs of translation sentences by using an extraction rule of the source language; extracting a target language segment from each target language sentence in the n pairs of translation sentences; and generating an extraction rule of the target language based on at least n target language segments extracted from n target language sentences. According to the solution provided in embodiments of this application, the extraction rule of the target language can be automatically generated based on the determined extraction rule of the source language.