基于知识融合的微信多媒体检索研究

WeChat Multimedia Retrieval Based on Knowledge Fusion

【目的/意义】随着微信中多媒体信息资源的增长,传统的微信检索已不能满足用户的检索需求,如何快速高效地从多媒体数据中检索出用户需要的信息是微信检索面临的关键问题。【方法/过程】为解决上述问题,本文提出一种基于知识融合的微信多媒体检索框架,利用知识融合技术来揭示不同模态数据之间潜在的关联。【结果/结论】基于知识融合的微信多媒体检索模型,将多媒体信息构成一个资源群体,可有效消除数据冗余,提高微信检索效率。

【Purpose/significance】With the growth of multimedia information resources in WeChat, the traditional WeChat search can no longer meet the user’s retrieval needs. How to quickly and efficiently retrieve the information users’ need from the multimedia data is the key issue of WeChat.【Method/process】In order to solve the above problems, this paper proposes a WeChat multimedia retrieval framework based on knowledge fusion, which uses knowledge fusion technology to reveal the potential correlation between different modal data.【Result/conclusion】The WeChat multimedia retrieval model based on knowledge fusion forms a resource group with multimedia information, which can effectively eliminate data redundancy and improve WeChat search efficiency.