Research On The Evaluation System Of Modern Tea Art By Artificial Intelligence

Tea art is a term that probably appeared in Taiwan in the 1970 s.Since ancient times,there were plenty of factual tea arts in China,but few of them had been refined and raised to the theoretical level.Under the background of the digital humanities,humanities scholars use more extensive data models for research.Narrative visualization is a branch in the field of digital humanities,which can be used to broaden the interpretation of works.By means of visualization,we can see the sequence of events on the timeline,and then make a high-level and abstract summary,which will help to discover the global model.Based on the narrative theory,this paper uses visual means such as emotional fluctuation line,theme river map,and narrative timeline to deconstruct 301 tea video works in three aspects: narrative sense,narrative content,and narrative timeline.The main conclusions are as follows:At the text entry stage,voice recognition applications such as Yunzhisheng,Xunfei,and Baidu were tried.The recognition rate of tea art video language was found to be low,far lower than its claimed recognition.Through the calculations of Tencent Wenzhi,Chinese Academy of Sciences nlpir,and Baidu sentiment analysis api,we have found seven types of emotional patterns,which can be used to explain the changes in emotional polarity along the narrative timeline in all materials.Respectively rise(R),accounting for 52.49%,fall-rise(FR),accounting for 26.58%,blank(B),accounting for 7.64%,rise-fall-rise(RFR),accounting for 9.97%,rise-fall(RF),accounting for 1.99% fall-rise-fall(FRF),accounting for 1% and zigzag(ZZ),accounting for 0.33%.The vast majority of tea art video ends with positive emotions.Although there are certain differences among the three algorithms,facial emotion recognition can be used in conjunction with natural language processing to perform sentiment analysis of tea art works.It is suggested that the thematic river map can be used as a better visualization method for narrative content.The theme river diagram can be close to the traditional affective analysis,which includes two aspects: emotional object and emotional polarity.There is no one way to show the absolute advantage of the color of river map.As far as tools are concerned,Tableau and Plartton's color matching,Coolors and Color Palette Generator's style photo extraction,and colorbrewer can be used as color matching tools.Although it is suggested that semantic analysis and graphic search can be combined,it is a good way to map the theme color.But judging from the results of the crowd survey,it may not work.This paper presents a narrative line which can be used to analyze the narrative theory of tea art.It is found that the existing tea art materials cover 4 kinds of narrative patterns.Flash back mode accounted for 50.83%,linear accounted for 31.56%,no time accounted for 21.59%,flashforward accounted for 1.66%.Because some of these works are not just a single narrative pattern,so the sum ratio is greater than 100%.The non-linear narrative accounts for nearly 70%.In addition,through the combination of narrative line and emotional line can be quantitative analysis of cadence,even the use of narrative line can be feedback to tea art text writing.