Unstructured Information Attribute Extraction Algorithm Based on Time Series and Deep Convolutional Neural Network

In order to automatically identify and extract valuable attribute information from a large amount of unstructured text data for further analysis, management, and utilization, a time series and deep convolutional neural network-based unstructured information attribute extraction algorithm is proposed. Taking enterprise union information as an example, a retrieval model for enterprise union information is constructed to retrieve enterprise union information. Convolutional neural network learning method is used to comprehensively extract the features of enterprise union information. Using time series methods to extract time related data from the characteristics of enterprise trade union information, and standardizing it into a unified date format, in order to better understand the internal structure of the data. By utilizing the powerful feature extraction capability of deep CNN, high-level attribute information closely related to time series features can be further extracted from preprocessed data through time series analysis, in order to more accurately capture key elements in enterprise union information. The experimental results show that the information attribute recommendation results of the proposed method are more accurate and efficient.