基于产业特征语义匹配模型的产业融合预测研究

Research on Industry Fusion Prediction Based on Industry Feature Semantic Matching Model

以BVD-Zephyr并购交易分析库为数据源,以漏斗式的三重筛选机制为框架,以Doc2vec语义相似度、LDA主题挖掘算法为基础构建基于产业特征的语义匹配模型,进行产业融合趋势预测,以智能可穿戴设备领域为例进行实证研究。语义匹配模型的关键在于设立契合度、贴合度两个指标以过滤、匹配的方式探索异常值,以减少噪音数据,提高产业融合趋势预测准确性。结果表明:产业融合沿产业链方向展开;电器设备领域以技术互补型的产业融合为主;医疗领域智能可穿戴的应用实践性需要深化(如智能反馈功能)。

Based on Doc2vec semantic similarity and LDA topic mining algorithm,this paper builds a semantic matching model based on industrial characteristics to predict the trend of industrial convergence.An empirical study is conducted in the field of smart wearable devices.The key of semantic matching model is to set two indexes of fit degree and fit degree to explore outliers by filtering and matching,so as to reduce noise data and improve the accuracy of industry convergence trend prediction.Results show that,industrial integration along the direction of the industrial chain,the field of electrical equipment is mainly based on the industrial integration of complementary technologies.Practical application of smart wearable in medical field needs to be deepened(such as intelligent feedback function).