Update ReCiter clustering so that it can be run locally and produce readable output about reciter

[jl987-Jie]

04-05-15 Update:

• Updated project name to ReCiter.
• Updated Lucene code to index title, journal, keyword field without tokenizing.
• Still need to add an untokenized affiliation field.
• Added a slf4j logger. Outputs the following metadata:
• article pmid
• article title
• article journal title
• article author names

Still need to work on outputting:

1. article author affiliation

Outputs precision and recall value for similarity thresholds from 0.1 to 0.9 in increments of 0.1.
Outputs the best similarity threshold out of these 9 values (0.1 to 0.9).
Outputs a CSV-like format to reciter.log file for each similarity threshold.

Note:

1. Make sure to have a local db connection (or change the db config in DbConnectionFactory.java) to enable ReCiter to retrieve the gold standard information.
2. To run the ReCiterExample.java for a specific person, please change the cwid, first name, middle name, last name in config.properties file in the project workspace.
3. The reciter.log file appends any log outputs, so to ensure a clean log, try to delete to file every time you run the ReCiterExample.java file for a particular cwid.

[jl987-Jie]

1. Beta version of CSV output finished.
2. Added AnalysisCSVWriter.java to output CSV output with file name csv_output.csv.
3. Added more options to the file config.properties.
   a). authorKeywords: keywords which relates to the author.
   b). similarityThreshold: cosine similarity
   c). coAuthors: co-authors of this author.
4. Running ReCiterExample.java will produce the csv file with configurations in config.properties.