Scientific Document Summarizer from the Web IR / NLP Group (WING), NUS
TODO: Caching json data based on input corpus. This is should be retrievable as well. Most scripts should be able to retrieve from the cache however this should only be uni directional. If the input corpus is cached, the script will retrieve data that is necessary for its processing. If not or if the right extent of data is not cached, then the pipeline needs to be manually run. (Should default pipeline be initiated?)