利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

19 条记录 (耗时 0.011 秒)

    ARTIFICIAL INTELLIGENCE SOCIAL FACILITATOR ENGINE
    2.
    发明公开
    ARTIFICIAL INTELLIGENCE SOCIAL FACILITATOR ENGINE 审中-公开

    公开(公告)号：EP4395242A1

    公开(公告)日：2024-07-03

    申请号：EP23218979.5

    申请日：2023-12-21

    申请人： Meta Platforms, Inc.

    发明人： WOODS, Philip ,  TSE, Daniel Gar-Shon ,  MUÑOZ, Alejandro Castillejo ,  AGRAWAL, Pravesh

    IPC分类号： H04L12/18 ,  H04L65/4038

    CPC分类号： H04L12/1822 ,  H04L12/1831 ,  G06F40/35 ,  H04L65/403 ,  H04L51/04

    摘要： Aspects of the present disclosure are directed to providing an artificial intelligence social facilitator engine (e.g., an "Al social bot") that can facilitate group conversations occurring over a network. The Al social bot can perform a series of logical steps in 3 phases: 1) a detect phase, 2) a decide phase, and 3) a deliver phase. In the detect phase, the Al social bot can look at many aspects of the conversation that already exist and try to find information that could be relevant to a decision about what to say. In the decide phase, the Al social bot can analyze the information and decide what type of conversational move to make. In the deliver phase, the Al social bot can use a large language model to formulate an actual text response corresponding to the information and strategy respectively formulated by the detect phase and the decide phase.

    VIDEOCONFERENCING MEETING SLOTS VIA SPECIFIC SECURE DEEP LINKS
    3.
    发明公开
    VIDEOCONFERENCING MEETING SLOTS VIA SPECIFIC SECURE DEEP LINKS 审中-公开

    公开(公告)号：EP4319053A1

    公开(公告)日：2024-02-07

    申请号：EP23189582.2

    申请日：2023-08-03

    申请人： TMRW Foundation IP SARL

    发明人： YERLI, Cevat

    IPC分类号： H04L9/40 ,  H04L12/18 ,  G06F21/62 ,  H04L61/301 ,  H04L65/4038 ,  H04L67/146 ,  H04L101/365

    摘要： A method for generating secure deep links comprises generating a deep link generation request; receiving a videoconferencing meeting slot list, wherein each videoconferencing meeting slot comprises at least a location within a videoconferencing space of a videoconferencing platform stored in memory; and generating a deep link that is unique for each videoconferencing meeting slot, the deep link encoding at least the location of the videoconferencing meeting slot within the videoconferencing space. Methods for generating spatial deep links for virtual spaces for inviting participants, and methods for generating distributed deep links using data fragments that may be reassembled, are also described.