Home - ILCogCompCuratorToHadoop/CuratorHadoopInterface GitHub Wiki
Welcome to the CuratorHadoopInterface wiki!
Project Pages
Documentation
- Project Readme: Start here to begin using the software.
- Final Project Report: A narrative summary of the 10-week project.
- Weekly Progress Reports: Week-by-week, what has been done and what we're working on.
- Project Roadmap: High-level, what has been done and what we're working on.
- Poster Presentation: Download a copy of our final poster.
Future Work
- Issues Running on the Altocumulus Cloud's Hadoop Cluster: Documents the problems we ran into when attempting to run the software on the Illinois Cloud Computing Testbed
- Curator Web Annotation Service: Describes the service that would allow web users to use our Curator server to process large numbers of documents and get the results back.
Solved Problems
Reference
Our Code
- Project Overview: Gives a visual overview of the interactions between existing code (like the Curator) and our tools here
- Infrastructure UML Diagram: Very important for people who want to modify the code for the Hadoop MapReduce job. Includes a text explanation and rationale for each of the classes responsible for setting up a job.
- Curator Reducer and Curator Client UML Diagram: Also very important for people who want to modify the code for the Hadoop MapReduce job. Documents the classes involved in actually getting Curator annotations on documents during a MapReduce job.
- Time Trials: A spreadsheet documenting the results of timed speedup tests.
Cognitive Computation Group's Tools
- Dependency Tree for Annotation Tools: Shows which how annotations depend on one another (for instance, that the chunker depends on the POS tagger and the tokenizer).
- Curator Annotation Identifiers and View Types: Shows how we convert between our enumerated type and the Curator's names for annotations. Also shows what "type" each annotation is (parse, cluster, or label).
- Curator Interface (Thrift)
- Curator Client Demo (Java)