Research On Knowledge Mining And Representation Of Product Family Appearance Design Based On Ontology

With the serious homogeneity of product performance in the current market,in order to consolidate the brand’s cognitive position in consumers,enterprises design family-based products,that is,according to the common characteristics of family products,the next-generation products are subject to a certain range of constraints.redesign.This requires the reintegration and expansion of the existing design experience and knowledge to form a unified specification,which can be applied to the next generation of product design.The innovative design thinking of the product comes from the designer’s design experience and the relevant data of the enterprise.In the early stage design,the designer needs to communicate effectively with the enterprise,so that the design knowledge can be accurately expressed.To revise the scheme,the two must communicate again,which increases the communication cost,leads to a long scheme design cycle and low product development efficiency.Therefore,based on the method of ontology representation,this research constructs an ontology conceptual model from the field of product family design,and realizes knowledge sharing and reuse in the product design process through the knowledge graph formed by the ontology model.First,analyze and study the elements of the product family design and the relationship between the elements.Collected and sorted out relevant information on Mazda cars as an example,mainly using web crawler technology to crawl online reviews.The collected corpus is reprocessed by natural language related processing methods,including word segmentation,stop word removal,and word frequency statistics.After data reprocessing,relevant calculation methods are used to determine key high-frequency words,combined with the characteristics of design knowledge,the target corpus of relevant elements of product design is limited,and the co-occurrence analysis method is used to study the co-occurrence of keywords..Second,construct a mining mechanism based on corpus extraction and clustering dimensionality reduction.Firstly,the word segmentation of the processed corpus text is analyzed by factor analysis and the common factors of knowledge concepts are extracted.Secondly,the key concepts are classified by the system clustering method,so as to construct the mining mechanism of product family design knowledge,so as to determine the knowledge concept.Then,the specific application of the method is explained from the three aspects of the correlation of key high-frequency words,main factors,and cluster analysis.Finally,the clustering genealogy diagram of the keywords is output,which is used to construct the semantics of the product family design knowledge ontology.Conceptual models provide data sources.Third,build a product family design knowledge graph in the form of ontology.By using the Protégé5.5 ontology modeling software,the knowledge concepts are defined in the attributes and categories in the software.Combined with the seven-step modeling method,an exterior design knowledge ontology model is constructed,taking Mazda series cars as an example,to realize the formal representation of exterior design knowledge ontology,so as to solve the problem of domain information heterogeneity.Through the Onto Graf plug-in in the Protégé 5.5 software,the ontology visualization of the Mazda series of vehicle exterior design knowledge is carried out to form an ontology knowledge map.The research takes Mazda series cars as examples,on the basis of the definition of the conceptual layer of the ontology model,the corpus content of design knowledge is determined through the relevant design elements of exterior design.Factor analysis and systematic clustering method are used to construct a knowledge mining mechanism for product family design knowledge based on corpus dimensionality reduction extraction and clustering,and key knowledge concepts with design elements are extracted from the acquired corpus texts.Finally,taking the Mazda series of cars as an example,the Protégé5.5 modeling software is used to carry out a more accurate abstract representation of the knowledge concept of Mazda’s exterior design,and build a knowledge graph in the form of ontology.The knowledge graph provides a certain foundation for the reuse and sharing of knowledge among different design systems,and provides a design reference for enterprises in product innovation design.