Academic literature recommendation method fusing title and abstract semantic relation
A semantic relationship and recommendation method technology, applied in neural learning methods, semantic analysis, natural language data processing, etc., can solve problems such as cold start and data sparsity, improve the quality of paper recommendation, alleviate sparsity and cold start problems Effect
Pending Publication Date: 2022-06-14
TAIYUAN UNIV OF TECH
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] In order to solve the problems of cold start and data sparsity in existing academic document recommendation methods, the present invention provides an academic document recommendation method that integrates the semantic relationship between titles and abstracts
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0029] The present invention will be further described below with reference to the accompanying drawings.
[0030] The invention proposes an academic document recommendation method integrating the semantic relationship of title and abstract. The concrete realization steps of this invention are as follows:
[0031] S100: Collect user-document interaction data, and perform data preprocessing on the dataset. The data set contains the documents that each user has interacted with in the history and the title abstracts of the corresponding documents. The interaction refers to whether the user has collected, browsed, or clicked on a document in the history.
[0032] S200: Build a document recommendation network that integrates the semantic relationship of title abstracts. The document recommendation network first obtains the vector representation of the words in the title abstract through a pre-trained BERT model, and then captures the semantic relationship between the title abstrac...
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
Login to View More Abstract
The invention belongs to the field of academic literature recommendation, and particularly relates to an academic literature recommendation method fusing title and abstract semantic relations. The academic literature recommendation method solves the problems of cold start and data sparsity of an existing academic literature recommendation method, and comprises the following steps of S100, collecting user-literature interaction data and performing data preprocessing; s200, building an academic literature recommendation network combining the text and the implicit feedback information; s300, inputting the preprocessed data set into a literature recommendation network, training the literature recommendation network by using a loss function, and storing a trained network model and parameters; and S400: using the trained network to calculate the preference score of each user for all other literatures which are not interacted, sorting the literatures according to the preference scores, and selecting the first N literatures to be recommended to the user. The academic literature recommendation effect is effectively ensured.
Description
technical field [0001] The invention belongs to the field of academic document recommendation, in particular to an academic document recommendation method integrating the semantic relationship between title and abstract. Background technique [0002] Academic literature refers to scholars who summarize and condense their research content, methods, experimental results and conclusions through articles after scientific research. It is a systematic elaboration and discussion on a certain issue. Great value. Scholars obtain scientific research information such as the latest research progress and research status in a certain discipline or field through academic literature, which stimulates the emergence of scientific research motivation and the emergence of academic inspiration. The traditional way of obtaining academic documents is to read paper documents. In the Internet age, document acquisition has become very easy. The Internet is flooded with a large number of academic do...
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
Login to View More IPC IPC(8): G06F40/258G06F40/30G06F16/9535G06N3/04G06N3/08
CPCG06F40/258G06F40/30G06F16/9535G06N3/04G06N3/08
Inventor 陈泽华陈雨民吕传建闫一帆
Owner TAIYUAN UNIV OF TECH