基于长短期记忆网络的心理健康数据分布式采集模型研究

Research on Distributed Collection Model of Mental Health Data Based on Long Short-Term Memory Networks

为了实现对心理健康数据的准确处理与分析,提出基于长短期记忆网络的心理健康数据分布式采集模型。利用长短期记忆网络的选择性记忆特性,对初始数据进行分解重构,将重构偏差较多的心理健康数据判定为冗余数据,同时将其过滤;将C/S架构作为模型的整体框架,使用数据源管理、数据采集管理、网络通信管理、数据缓存管理四个板块完成分布式采集任务,利用长连接、变长数据包和缓存机制确保数据传输时效,运用线程安全队列来维护模型操作安全。经仿真分析可知:该模型具有较高的数据分布式采集精度,采集速率快且稳定性强。

In order to realize accurate processing and analysis of mental health data,this paper proposes a distributed mental health data collection model based on long short-term memory networks.The initial data are decomposed and reconstructed by using the selective memory characteristics of the long short-term memory networks.The mental health data with many reconstruction deviations are judged as redundant data,and then filtered.The C/S architecture is taken as the overall framework of the model,and the distributed collection task is completed by using four sections:data source management,data collection management,network communication management and data cache management.The long connection,variable length packet and cache mechanism are used to ensure the time of data transmission,and the thread-safe queue is used to maintain the operation safety of the model.Simulation results show that the model has high precision of distributed data acquisition,fast data acquisition rate and strong stability.