基于word2vec的媒体平台和学术平台移动健康比较研究

Mobile health refers to the use of mobile communication technology to deliver healthcare services and information. This study compares mobile health between media platforms and academic platforms, with the aim of facilitating sustainable development of mobile health in China and beyond. Word2vec is a neural network word embedding algorithm that can effectively measure word similarity at the syntactic and semantic levels.This study collects data from the media platform of CN-HEALTHCARE and the academic platform of China National Knowledge Infrastructure (CNKI) separately. The abstracts of all the literature regarding mobile health on CNKI and all the media reports regarding mobile health on CN-HEALTHCARE were separately used as the training corpus. Two word2vec models were trained based on the collected corpus. Then, the high-frequency words in the two corpora were compared and analyzed. The top 20 similar words of mobile health in the two word2vec models were obtained and compared to find out the differences between media platforms and academic platforms in terms of their focus on mobile health. Following this, dimensionality of the word vectors in the two word2vec models were reduced. Visualization of the word embedding of the whole words was conducted and the difference was compared. The results of this study show the differences of mobile health between media platforms and academic platforms, providing a new perspective for mobile health research.