| dc.description.abstract |
Flat website design has become more and more popular in recent years due to its simplicity and elegance. Nevertheless, the current flat design thinking is mainly applied to conventional websites such as corporate websites. Recently, the increase in public demands has significantly increased the browsing frequency of government websites. To optimize the government website, this study establishes a government website design framework from the perspective of flattening. The established framework can extract flat design elements from the massive flat website design schemes in the past. Based on flat design elements, the particle swarm algorithm (PSO) algorithm in the established framework realizes the design and optimization of government websites. Subsequently, the performance of the CNN, which is used as the fitness function of the PSO algorithm, in predicting public acceptance is analyzed with a case study. The case results show that the average relative errors between the prediction results predicted by CNN and the real public acceptance in two provinces are 2.5416% and 1.4788%, respectively, which indicates that the predicted and real results are in good agreement. Moreover, the linear correlation coefficients between predicted and real public acceptance are 0.9553 and 0.9937, respectively, which further indicates that CNN is reliable as a fitness function of the PSO algorithm. Therefore, it is feasible to use the PSO algorithm to optimize the government website design scheme. © 2022 Wenyan Zhong and Ming Liu. |
|