Abstract: An electronic device includes a memory configured to store instructions and a processor electrically connected to the memory and configured to execute the instructions, in which when the instructions are executed by the processor, the processor is configured to perform a plurality of operations, in which the plurality of operations includes deriving a frequently-asked-questions (FAQ) pair from speech data based on a neural network model trained in an end-to-end manner, in which the neural network model is based on a multi-modal language model (LM) capable of using text data and speech data simultaneously, and contrastive learning is performed on the neural network model based on symmetric loss to shift speech data, which is original data, to text data, which is augmented data

Abstract: A system for a vehicle is provided. The system may include a wireless interface configured to connect a server with an input device and an output device of the vehicle. The server may store sample data, associated with the vehicle, that match a plurality of output responses corresponding respectively to the sample data. The server may generate, based on input data received, via the wireless interface from the input device of the vehicle, a sample datum from the stored sample data, retrieve, from the memory, an output response, of the plurality of output response, that matches the sample datum, output, via the wireless interface to the output device of the vehicle, the retrieved output response, perform multi-task learning based on the input data and the sample data, identify, based on the retrieved output response, a user intention, and cause, based on the identified user intention, the vehicle to be controlled.

Abstract: A method of training a part-of-speech (POS) tagging model includes: separating an input sentence into units of syllables to generate an input sequence; encoding, using at least one encoder included in a part-of-speech (POS) tagging model, the input sequence; generating, based on the encoded input sequence and using a first discriminator included in the POS tagging model, a POS tagging result; and generating, based on the encoded input sequence and using a second discriminator included in the POS tagging model, a spacing result.

Abstract: The present invention relates to a perovskite solar cell module and a manufacturing method for same. The perovskite solar cell module comprises a plurality of perovskite solar cells disposed on a substrate, each of the perovskite solar cells comprising: a first electrode, a first charge transport layer on the first electrode, an optical active layer formed of a perovskite crystal structure, and a second charge transport layer, which are laminated in this order; and a second electrode laminated on the second charge transport layer, wherein the second electrode included in each of the cells can be electrically connected in series to the first electrode of the closest perovskite solar cell and enhance the photoelectric conversion efficiency of the perovskite solar cell module.

Abstract: Disclosed herein is a system including an answer determination module configured to analyze an input sentence to determine whether to answer the input sentence, a learning module configured to output a domain corresponding to the input sentence and a plurality of categories to which the input sentence belongs when it is determined to answer the input sentence, and an output module configured to output an answer to the input sentence, wherein the learning module performs multi-task learning using the input sentence as input data and using as output data the domain corresponding to the input sentence and the plurality of categories to which the input sentence belongs.

Abstract: Provided is a named entity recognition system, including: an input module configured to recognize a speech input of a user and convert the speech input into text; a preprocessing module configured to separate the text in units of syllables and perform transformation; and a learning module configured to perform multi-task learning for recognizing a named entity and identifying a boundary of spacing with respect to the transformed text, and output a result of recognizing the named entity and a result of identifying the boundary of spacing, based on recognizing the named entity and identifying the boundary of spacing.

Abstract: As a speech act analysis device, the speech act analysis device includes: a word similarity calculator that receives an input utterance vector that is vectorized from information on at least one or more words forming an input utterance, and a previous speech act vector that is vectorized from speech act information with respect to a previous utterance of the input utterance, and generates an input utterance similarity vector that reflects similarity between the input utterance vector and the previous speech act vector; a conversation vector generator that generates a conversation unit input utterance vector that is vectorized from information with respect to the input utterance in a conversation including the input utterance by inputting the input utterance similarity vector in a convolution neural network; a conversation similarity calculator that receives a speaker vector that is vectorized from speaker information of the input utterance, and generates a conversation unit input utterance similarity vector t