| With the advent of the era of knowledge economy and the Internet technology develop at top speed,the traditional closed innovation model cannot meet the needs of the fierce competition of enterprises,and enterprises are striving to seek new innovation models.Crowdsourcing design mode has been proved to be an effective mode to obtain knowledge resources from external groups to achieve innovation goals.Through this model,enterprises can obtain creative knowledge from public users,make up for the lack of internal innovation resources of the enterprise,effectively reduce the innovation cost of the enterprise,and enhance the innovation performance of the enterprise.Knowledge as an intangible resource is gradually becoming a key factor in promoting social progress,and the opening and sharing of knowledge has become an urgent need for people.Crowdsourcing design text of small household appliances is taken as the research object of this paper,and researches on knowledge modeling,knowledge recognition and knowledge sharing are carried out,and a knowledge sharing platform for crowdsourcing design texts of small household appliances is developed.First,the content related to crowdsourcing design of small household appliances was analyzed,and the texts related to crowdsourcing design of small household appliances were collected and processed.Combined with the characteristics of crowdsourcing design of small household appliances,a corpus in the field of crowdsourcing design of small household appliances is established by using natural language processing related technologies.The Bi LSTM-CRF algorithm model is constructed and applied in the field of crowdsourcing design of small household appliances,which effectively extracts relevant entities in the field of crowdsourcing design of small household appliances.Secondly,an entity relation extraction method based on meta-knowledge and dependency model is proposed.Input the sentence into the syntactic dependency analysis model,construct the candidate triplet data and entity chain data,extract the triplet with the candidate triplet data through the generalization matching of the seed template,and use the meta-knowledge template to process the entity chain data to supplement the triplet.The knowledge graph is constructed and displayed based on entity-relation-entity triples,and the knowledge graph is completed by using the graph attention neural network to make the knowledge graph more complete.Finally,the text knowledge sharing platform of crowdsourcing design of small household appliances is analyzed,and the knowledge sharing platform of crowdsourcing design text of small household appliances is designed and developed.The platform has the functions of text entity recognition,entity relationship extraction,and knowledge graph display functions for crowdsourcing design of small household appliances,etc.which can help users quickly understand the knowledge content of documents related to the design of small household appliances,and carry out related applications. |