Abstract:
With the continuous development of Internet, cloud computing, 5g technology and we media industry, people generate and obtain more and more information every day, and mos...Show MoreMetadata
Abstract:
With the continuous development of Internet, cloud computing, 5g technology and we media industry, people generate and obtain more and more information every day, and most of these information exists in the form of text. Automatic text technology can quickly help people obtain clear and readable information in massive text. Aiming at the problem that the traditional text technology based on deep learning can not generate high-quality long text summary, this paper explores a deep proxy text generation technology based on reinforcement learning. The model divides the long-term and short-term memory network into different agents, and each agent is responsible for encoding a part of the source text. After that, each agent will broadcast its own coding results, so that all agents can share the global context information. After the final coding, the model integrates the information from multiple agents through a context agent attention mechanism and transmits it to the decoder, so as to optimize the process of long text summary generation.
Published in: 2022 International Conference on Artificial Intelligence of Things and Crowdsensing (AIoTCs)
Date of Conference: 26-28 October 2022
Date Added to IEEE Xplore: 20 April 2023
ISBN Information: