Commodity classification method and device
A classification method and product technology, applied in text database clustering/classification, unstructured text data retrieval, etc., can solve problems such as inaccurate product classification, and achieve the effect of good accuracy and accurate product classification
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0053] figure 1 A schematic diagram of a commodity classification process provided by an embodiment of the present invention, the process includes the following steps:
[0054] S101: Identify the taxpayer enterprise information and invoiced product information in the invoice.
[0055] The product classification method provided by the embodiment of the present invention is applied to electronic devices with data analysis and machine learning capabilities.
[0056] When electronic equipment determines the type of goods, it needs to identify the taxpayer enterprise information and invoiced commodity information in the invoice. Generally, the taxpayer enterprise information is mostly continuous and standardized information, and the invoiced commodity information is mostly non-standard information. For example, the taxpayer enterprise information may be enterprise organization code information, enterprise name information, etc., and the invoiced product information may be the prod...
Embodiment 2
[0072] In order to reduce the amount of calculation in the process of product classification, on the basis of the above-mentioned embodiments, in the embodiment of the present invention, after determining each first participle of the taxpayer enterprise information, based on the word-to-word vector algorithm, determine Before each first vector corresponding to each first participle, the method also includes:
[0073] For every adjacent two first participle, judge whether to comprise the third participle after this adjacent two first participle combination in the dictionary storehouse saved in advance, if yes, adopt described third participle to pair adjacent two first participle A participle is updated.
[0074] A dictionary database is pre-stored in the electronic device, and the dictionary database contains word segmentation information. After determining each first participle of taxpayer enterprise information, for every two adjacent first participles, judge whether the th...
Embodiment 3
[0078] On the basis of the above embodiments, in the embodiment of the present invention, the CNN includes a convolutional layer and a maximum pooling layer; wherein the convolutional layer includes at least one convolution kernel;
[0079] According to the weight coefficient in the CNN, performing convolution calculation on the first vector, and determining the first feature vector corresponding to the first vector includes:
[0080] For each convolution kernel in the convolution layer, according to the weight coefficient corresponding to the convolution kernel, the convolution calculation is performed on the first vector, and the third feature vector corresponding to the first vector is determined; for each Each third feature vector determined by the convolution kernel is input to the maximum pooling layer, and the first feature vector corresponding to the first vector is determined from each third feature vector.
[0081] The CNN includes a convolution layer, and the convol...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More