Research on the Key Techniques of Semantic Mining of Information Digest in the Field of Agricultural Major Crops Based on Deep Learning

Abstract : Nowadays application scopes of deep learning research in the machine learning subfield have been gradually expanded, mainly in the field of computer vision and natural language processing. However, in the latter NLP field, there is very little semantic excavation research on agricultural literature data. This paper bases on the attempting to combine relevant paradigms of semantic mining techniques and characteristics of agricultural digest data, for the service of providing new methods and technologies of information acquisition and analysis in the agricultural information domain. Data cleaning methods and data mining experiment are mainly based on deep learning algorithms, which are Seq2Seq and attention mechanism. Finally, through qualitative evaluation and quantitative evaluation of the experimental results, which based on the ROUGE evaluation index system, the experiment shows that the semantic mining model has reached the optimal level of model evaluation in the certain range.
Document type :
Conference papers
Complete list of metadatas

Cited literature [8 references]  Display  Hide  Download

https://hal.inria.fr/hal-02111530
Contributor : Hal Ifip <>
Submitted on : Friday, April 26, 2019 - 9:59:03 AM
Last modification on : Friday, April 26, 2019 - 10:47:30 AM

File

 Restricted access
To satisfy the distribution rights of the publisher, the document is embargoed until : 2022-01-01

Please log in to resquest access to the document

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Hao Gong, Yunpeng Cui, Ping Qian. Research on the Key Techniques of Semantic Mining of Information Digest in the Field of Agricultural Major Crops Based on Deep Learning. 11th International Conference on Computer and Computing Technologies in Agriculture (CCTA), Aug 2017, Jilin, China. pp.496-507, ⟨10.1007/978-3-030-06179-1_49⟩. ⟨hal-02111530⟩

Share

Metrics

Record views

10