Research And Application Of Video Feature Recommendation Method In Small And Micro Culture Enterprises

Today,with the vigorous development of the cultural industry,small and micro cultural enterprises can not be accomplished.They constitute not only the main component of the cultural industry,but also the inexhaustible motive force of cultural innovation,which provides a huge channel for the job market.This subject comes from the Hualu new media operation platform,platform mainly completes China recorded with video content providers and operators of video copyright and revenue management.China records the rights of the video content provider and sends the video to the operator to get information about the use of the video.At present,the vast number of small and micro enterprises represented by China record market research,statistics,user demand rate,ratings analysis and other ways to obtain audience demand.These methods consume a lot of manpower and material resources,and the results depend solely on personal experience.Without scientific and effective indicators,the conclusions obtained are sometimes not scientific enough.Therefore,we need to study data mining,do related data analysis,and guide the production and purchase of small and micro enterprise video in a quantitative way.According to the culture of small and micro enterprises to do video recommendation needs,combined with information entropy theory to improve the formation of adaptive genetic algorithm based on information entropy of basic genetic algorithm(IAGA algorithm)is used to optimize the initial weights of BP neural network and IAGABP threshold algorithm,the algorithm is applied to video,video to use for input video features for IAGABP video output form features recommended model,the model is applied to the small and micro enterprise culture characteristic of video recommendation system.The main contents and work are as follows:(1)Complete video preprocessing,video processing and video data processing.(2)According to the characteristics of video recommendation needs,research and analysis of the genetic algorithm and BP neural network,and puts forward the improved algorithm:adaptive genetic algorithm based on information entropy(IAGA)to optimize the initial weights of BP neural network and IAGABP algorithm to form the threshold.The formula for calculating the change of cross probability and mutation probability,and calculates the population entropy of each generation,the population entropy as the crossover probability and mutation probability formula of the independent variable,to control the genetic population entropy change trend,thus forming the adaptive genetic algorithm based on information entropy(IAGA algorithm),the improved algorithm for to optimize the initial weights of BP neural network formation threshold IAGABP.Establish IAGABP video feature recommendation model:video use as input,video features as output.(3)Design and implement a video feature recommendation system for small and micro enterprises.In the system,the IAGABP algorithm is used to analyze the video usage and the data related to video features to predict the video feature distribution.And the results are displayed in tabular form and graphic form,which provide a more scientific basis for small and micro cultural enterprises to purchase or make video selections.