LVTIA: A new method for keyphrase extraction from scientific video lectures

作者:

Highlights:

• LVTIA is a new method for video lecture indexing using statistical features.

• The text extracted from audio and video frames are considered to extract keyphrases.

• LVTIA uses global and local features for candidate phrases, using video time intervals.

• The proposed algorithm is evaluated based on precision, recall, f-measure and MAP@K.

• LVTIA is tested on five datasets with different languages (English and Persian).

摘要

•LVTIA is a new method for video lecture indexing using statistical features.•The text extracted from audio and video frames are considered to extract keyphrases.•LVTIA uses global and local features for candidate phrases, using video time intervals.•The proposed algorithm is evaluated based on precision, recall, f-measure and MAP@K.•LVTIA is tested on five datasets with different languages (English and Persian).

论文关键词:Multimedia indexing,Video lecture indexing,Text mining,Keyword extraction,Keyphrase extraction

论文评审过程:Received 2 July 2021, Revised 10 September 2021, Accepted 17 October 2021, Available online 17 November 2021, Version of Record 17 November 2021.

论文官网地址:https://doi.org/10.1016/j.ipm.2021.102802