SimpleNLG-ZH is a realisation engine for Mandarin that follows the software design paradigm of SimpleNLG, i.e., keeping a clear separation between morphological and syntactic operations. Mandarin, as a highly analytical language, needs far fewer morphological operations but many more syntactic constraints than English. SimpleNLG-ZH (“Zhongwen” is Mandarin for “Chinese'”) was first built as a realiser for generating referring expressions in Mandarin which are mostly noun phrases together with simple verb phrases, and then extended to cover other constructions and phenomena in Mandarin. The current version of SimpleNLG-ZH was developed as an adaptation from V4.4.8 of the original SimpleNLG-EN.
Chen G., van Deemter K., and Lin C. SimpleNLG-ZH: a Linguistic Realisation Engine for Mandarin, The 11th International Conference on Natural Language Generation (INLG), Tilburg, Netherlands, 2018.
Joint Sentiment-Topic (JST) Model
The joint sentiment-topic (JST) model is a weakly-supervised hierarchical Bayesian model for detecting document-level sentiment and extracting sentiment-bearing topics from text simultaneously. The only supervision used for JST learning is a small set of domain-independent sentiment clues; no labelled documents are used.
The JST source code (written in C++) can be downloaded here.
Lin, C., He, Y., Everson, R. and Rueger, S. Weakly-Supervised Joint Sentiment-Topic Detection from Text, IEEE Transactions on Knowledge and Data Engineering (TDKE), 2011.
Lin, C. and He, Y. Joint Sentiment/Topic Model for Sentiment Analysis, The 18th ACM Conference on Information and Knowledge Management (CIKM), Hong Kong, China, 2009.