摘要
随着网络多媒体技术的飞速发展,网络视频在各大媒体平台上开始呈现爆炸式增长。面对数量如此之多的视频资源,利用传统关键词匹配来进行视频检索已经无法满足部分平台用户的日常需求。为了更好地为平台用户提供高效准确的视频检索,本文融合异构信息网络和图嵌入技术,通过自然语言处理一些新的模型,从用户输入的一段文本信息进行数据分析,与视频数据库中视频简介信息进行相似度匹配,从而更加个性化地为用户提供视频检索推荐。实验证明,本文提出的模型在NDCG评价指标下得到0.6557的分数,取得了较好的实验效果。
With the continuous development of Internet multimedia technology,the number of network videos in major media platforms has exploded.Faced with so many video resources,the use of traditional keywords for video retrieval has been unable to meet the daily needs of some platform users.In order to better for users to provide efficient and accurate video retrieval platform,in this paper,the fusion of heterogeneous information network and graph embedding technology,through the natural language processing some of the new model,a piece of text can be input from the user information for data analysis,with the introduction of video information in the video database similarity matching,thus more personalized to provide users with video is recommended.The experimental results show that the model proposed in this paper can get the score of 0.6557 under NDCG evaluation index and achieve good results.
作者
杨舰
YANG Jian(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China)
出处
《电视技术》
2022年第7期47-50,共4页
Video Engineering
关键词
匹配
异构信息网络
图嵌入
视频检索推荐
keyword matching
heterogeneous information network
embedded figure
video retrieval recommendation