Apsect Based Sentiment Analysis (ABSA) python implementations
- SemEval 2014 solutions:
- NRC Canada: http://www.aclweb.org/anthology/S14-2076
- JUSCE
- SemEval 2015 solutions: TBD
- Deep Learning solutions for ABSA:
- CNN based impl by Wang & Liu (Stanford): https://cs224d.stanford.edu/reports/WangBo.pdf
- LSTM based impl by Tang et. al.: https://arxiv.org/pdf/1605.08900.pdf
For stanford corenlp python wrapper, https://github.com/dasmith/stanford-corenlp-python is used
- Need to run 'python corenlp.py' first. This starts a RPC service that is invoked by other python wrapped APIs.
- Since this RPC invoations, CANNOT RUN PARALLEL requests!
- Need to make sure 'jsonrpc.py' is in the same folder as 'corenlp.py'
- import jsonrpc, that's it!
reference for CRF:
- http://sklearn-crfsuite.readthedocs.io/en/latest/tutorial.html
- http://stats.stackexchange.com/questions/38216/implementation-of-crf-in-python
reference for NLKT parse tree to NX graph:
- http://stackoverflow.com/questions/25815002/nltk-tree-data-structure-finding-a-node-its-parent-or-children
- http://stackoverflow.com/questions/29397460/extract-parent-and-child-node-from-python-tree
- http://www.nltk.org/api/nltk.html#nltk.tree.ParentedTree
Aspect Location (May be TBD):
-
clustering paper: http://www.aclweb.org/anthology/N/N16/N16-1093.pdf