Probabilistic topic modeling in multilingual settings: An overview of its methodology and applications

作者:

Highlights:

• A systematic overview of multilingual probabilistic topic modeling (MuPTM).

• A tutorial on methodology, modeling, training, output, inference and evaluation of MuPTM.

• Language-independent and language-pair independent data representations.

• A model-independent framework and applications in various cross-lingual tasks.

• A complete MuPTM-based framework for cross-lingual semantic similarity.

摘要

•A systematic overview of multilingual probabilistic topic modeling (MuPTM).•A tutorial on methodology, modeling, training, output, inference and evaluation of MuPTM.•Language-independent and language-pair independent data representations.•A model-independent framework and applications in various cross-lingual tasks.•A complete MuPTM-based framework for cross-lingual semantic similarity.

论文关键词:Multilingual probabilistic topic models,Cross-lingual text mining,Cross-lingual knowledge transfer,Cross-lingual information retrieval,Language-independent data representation,Non-parallel data

论文评审过程:Received 4 February 2013, Revised 11 August 2014, Accepted 18 August 2014, Available online 7 October 2014.

论文官网地址:https://doi.org/10.1016/j.ipm.2014.08.003