Research On Intelligent Layout Of Digital Newspapers For Regional Media

Newspaper pages have a reasonable and beautiful arrangement of news,which allows readers to quickly obtain valuable information and gain pleasure.This is the feature of newspaper in the mass media.However,for typesetters,generating an intuitive,readable and beautiful newspaper layout is a time-consuming manual task.When readers are faced with multiple newspapers,they usually only pick the one with the most attractive layout to read.But assessing the aesthetics and attractiveness of a newspaper’s layout is a very subjective task.It is very challenging to get a computer to make an objective assessment of the aesthetic quality of a newspaper’s layout.This paper focuses on the design and implementation of the automatic generation method of digital newspaper layout and the evaluation method of digital newspaper layout.In addition,an online intelligent layout system is designed.The innovative work results of this paper mainly include:Combining Bayesian network and constrained programming technology,this paper proposed an automatic generation method of digital newspaper layout.Firstly,based on the historical layout and expert experience,we learned and inferred the structure and key attributes of the digital newspaper layout,so that the newly generated layout has a history style;Then a mixed integer constrained programming model was proposed by the inference results,which significantly reduces the solution space of the model and improves the layout quality.Our inference model provides a variety of available candidate structures and the programming model ensures that the news in the newspaper layout does not overlap,does not overflow,and has good alignment performance.This thesis proposes an aesthetic evaluation method for the layout of digital newspapers,which combines deep learning technology with the construction of layout indicators to objectively evaluate the quality of newspaper layout.First,a dataset on the aesthetic quality of layout is introduced.The dataset includes 3429 newspaper layout images and corresponding quality levels.Then,based on the improvement of the target detection algorithm YOLOv3,an image aesthetic evaluation model is designed,which integrates the overall features of the layout image and the layout features between the news to conduct aesthetic prediction of the layout quality.In addition,based on expert experience and news layout characteristics,objective evaluation indicators for layout coordination,balance,alignment,news distribution,and layout whitespace have been designed.Finally,the layout is graded based on aesthetic prediction results and objective evaluation indicators.A digital newspaper online typesetting system is designed.The system modularizes layout generation and layout evaluation.During the layout process,users can choose to generate a layout that will serve as an excellent layout to adjust and enrich the collection of historical excellent layouts,to indirectly modify the classification model in the layout generation component,thereby generating personalized layouts based on user preferences.In addition,the system will record the user’s choices and preferences to adjust the scoring weight of the layout scoring component.The system integrates user behavior into the system iteration process,making the system more intelligent.