Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A content-related advertisement delivery method and system based on bi-lstm-crf model

A technology for advertising placement and models, applied in the direction of neural learning methods, biological neural network models, instruments, etc., to improve the recognition effect and improve the effect of precise placement

Active Publication Date: 2022-06-14
HARBIN INST OF TECH
View PDF21 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of the above problems, the present invention proposes a content-related advertisement delivery method and system based on the Bi-LSTM-CRF model to solve the problem of small-scale data sets based on The named entity recognition model of deep learning, because it is not easy to automatically obtain features, makes it difficult for the model to achieve a good recognition effect, which further leads to the problem of inaccurate delivery of advertising recommendations

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A content-related advertisement delivery method and system based on bi-lstm-crf model
  • A content-related advertisement delivery method and system based on bi-lstm-crf model
  • A content-related advertisement delivery method and system based on bi-lstm-crf model

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment 1

Specific embodiment one

[0070] Verify the effectiveness of the method of the present invention.

[0071] The experiment uses the post data obtained from the second-hand community. Through manual labeling, the data set contains 19449 post data, and there are a total of 29859 commodity entities after labeling. The experiment is performed by a computer with 2 core CPU and 8G memory, using pytorch The framework implements the algorithm.

[0072] Divide the marked corpus into training set, verification set and test set according to the ratio of 8:1:1 for model training. To find the best parameter settings for the model, a parameter search method is employed. In this method, the word vector dimension is set between [200, 256, 300], the number of units in the LSTM layer is set between [64, 128], and the value of dropout is between [0.4, 0.5, 0.6]. The optimal parameter combination of the model obtained from the final test is shown in Table 6.

[0073] Table 6 Model optimal train...

specific Embodiment 2

[0079] From the analysis of the experimental data in Table 7, it can be seen that the accuracy rate of model 15 is 0.05% lower than that of the baseline model (namely model 3), but its recall rate is 4.15% higher, and the F1 value is 2.31% higher. The recognition effect is among all best in model. Integrating the influence of different models after fusing different feature combinations, the experimental data is drawn as follows: Figure 4As shown, from the perspective of recall rate and F1 value, the Bi-LSTM-CRF model with multi-feature fusion is better. Compared with the experimental results of the baseline model (ie model 3), the recall rate has increased by up to 4.15%, and the F1 The highest value is increased by 2.31%, which shows that the extra features proposed by the present invention are effective in combination with the characteristics of the entity itself. These feature combinations have improved the recognition quality of named entities to a certain extent. The mul...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A content-related advertising delivery method and system based on the Bi-LSTM-CRF model, belonging to the technical field of advertising, used to solve the named entity recognition model based on deep learning for small-scale data sets, because it is not easy to automatically obtain features, As a result, it is difficult for the model to achieve a good recognition effect, which further leads to the problem that the advertisement recommendation cannot be placed accurately. The technical points of the present invention include: inputting the training data set into the Bi-LSTM-CRF model for training to obtain the optimal prediction model; inputting the data to be predicted into the optimal prediction model to obtain the predicted commodity words; Advertisement, obtain the advertisement information with the highest matching degree; deliver the advertisement carrying the advertisement information. Based on the Bi-LSTM-CRF algorithm, the present invention combines the characteristics of commodity words, enhances data through feature engineering, makes the data have richer semantics, and constructs a system suitable for document commodity word extraction for content-related advertisements Recommendations improve the effect of accurate advertising.

Description

technical field [0001] The invention relates to the technical field of advertisement delivery, in particular to a content-related advertisement delivery method and system based on a Bi-LSTM-CRF model. Background technique [0002] The second-hand community is a network platform for users to publish and browse posts. Users can post posts to express their demands for idle transfers or browse posts to see if they meet their purchase intentions. For the second-hand community, the main way of income is advertising. Since the advertising process of the second-hand community needs to be accurately placed, it is necessary to start from the content of the community to analyze the user's purchase intention and use this to place advertisements, that is, content-related advertisements. The main idea of ​​contextual advertising is to deliver advertisements related to the content of the webpage while providing users with the content of the webpage. In the prior art, a method of keyword e...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06Q30/02G06F40/295G06N3/04G06N3/08
CPCG06Q30/0244G06F40/295G06N3/049G06N3/08G06N3/045
Inventor 景东张大勇卓兴良
Owner HARBIN INST OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com