Title:
VISUAL OBJECT GUIDANCE-BASED SOCIAL MEDIA SHORT TEXT NAMED ENTITY IDENTIFICATION METHOD
Document Type and Number:
WIPO Patent Application WO/2021/135193
Kind Code:
A1
Abstract:
A visual object guidance-based social media short text named entity identification method, comprising: establishing a text description data set of social media short text, and providing visual image data related to text description; performing object detection on the visual image data, and performing visual word semantic extraction on a detected visual object category; carrying out semantic extraction on the original text description, and carrying out text feature extraction on text semantics; calculating similarity between visual word semantics and text features, and performing weighted summation on the visual word semantics; calculating the correlation between multi-modal semantics and original text features to obtain a final word semantic vector; and inputting the final word semantic vector into a conditional random field to identify an entity boundary and an entity category, an obtained result being an identified named entity. According to the method, the entity type can be determined by utilizing visual object information, and social media short text data lacking semantic information is supplemented.
Inventors:
CAI YI (CN)
ZHENG CHANGMENG (CN)
ZHENG CHANGMENG (CN)
Application Number:
PCT/CN2020/105074
Publication Date:
July 08, 2021
Filing Date:
July 28, 2020
Export Citation:
Assignee:
UNIV SOUTH CHINA TECH (CN)
International Classes:
G06F40/295
Foreign References:
| CN111126069A | 2020-05-08 | |||
| CN108628823A | 2018-10-09 | |||
| CN110334357A | 2019-10-15 | |||
| CN109522553A | 2019-03-26 | |||
| US20190354582A1 | 2019-11-21 |
Attorney, Agent or Firm:
ETR LAW FIRM (CN)
Download PDF: