基于词向量和远程监督学习的格萨尔领域知识发现方法研究

Knowledge Discovery in Gesar Studies Based on Word Vector and Distant Learning

当前信息和知识在科技成果推动下的急速增长,与人们高效进行知识发现的需求之间的矛盾日渐加深,利用人工智能等技术提升知识发现效率是缓解这一矛盾的可试之法。文章从通用领域与格萨尔学科领域采集多元数据,基于词向量技术完成对数据的统一表示与向量关系计算,再基于远程监督学习技术利用知识图谱既有数据完成对实验数据的弱监督关系抽取,进而对人物关系进行预测,最终实现格萨尔学科领域中的人物关系知识发现,在方法上是对传统基于规则的知识发现方法的一种补充。

At present,the contradiction between the rapid growth of knowledge and information driven by scientific and technological achievements and the need for people to efficiently discover knowledge is deepening.Using technology such as artificial intelligence to improve the efficiency of knowledge discovery is a testable method to alleviate this contradiction.The article collects multivariate data from the general field and in Gesar studies,and completes the unified representation and vector relationship calculation based on word vector technology.Then,based on the distant learning technology,the extraction of the weak supervised relationship of experimental data has been done by using the data of knowledge graph so as to realize the knowledge discovery of personal relationship in Gesar studies by predicting.This method complements the traditional rule-based approaches of knowledge discovery.