NAME
    UMLS::SenseRelate CHANGES

  Changes from version 0.27 to 0.29
    1. added three additional aggregators (andness, disp and closeness)

  Changes from version 0.25 to 0.27
    1. added two additional aggregators (orness and max)

  Changes from version 0.23 to 0.25
    1. added the --info option to the umls-senserelate-evaluation.pl program
    to print out the accuracy of the individual target words.

    2. add the umls-senserelate-ttest.pl program which takes two info files
    generated by the umls-senserelate-evaluation.pl program and calculates
    their signficance using the Statistics::TTest cpan module.

  Changes from version 0.21 to 0.23
    1. added an --intrinsic [santos|sanchez] option to use the information
    content measures with the intrinisic information content metrics
    proposed by Santos, et al 2004 and Sanchez, et al 2011. This is
    currently only available in the umls-targetword-senserelate.pl program
    but I will be adding it to the allwords once I get it sped up a bit.

  Changes from version 0.19 to 0.21
    1. modified the --weight option so that stopwords are included in the
    distance

  Changes from version 0.17 to 0.19
    1. added --weight option to weight the cuis/terms/words based on the
    distance they are from the target word; this is in allwords as well as
    target word

  Changes from version 0.15 to 0.17
    1. added a --cuis option to use the surrounding CUIs obtained by MetaMap
    when using the --mmxml formatted input file

    2. speed things up - working more on this

    3. added a --loadcache and --getcache option to target word programs

  Changes from version 0.13 to 0.15
    1. added a --restrict option which uses only the terms that have a
    mapping to a UMLS concept in the window - similar to --stoplist only
    using the mapping to determine if the term is a content term or not

    2. modified the --precision option to be used with the similarity
    measures

  Changes from version 0.11 to 0.13
    1. add the option to use the default parameter files for the vector and
    the information content files

    2. added a aw-xml check to the all words disambiguation utils/ program.

    3. added the --mmxml option to the target words disambiguation utils/
    program

    4. added the --candidate option to the target words disambiguation
    utils/ program

    5. added format checks to teh target words disambiguation utils/ program

  Changes from version 0.09 to 0.11
    1. added time stamp to the default log directory in
    umls-targetword-senserelate.pl - forgot that one

    2. updated the stderr ouput parameters in the utils programs

    3. added XML::Twig dependency

    4. added SYNOPSIS to AllWords.pm module

    5. changed internal modules to account for changes in the output
    containers in the UMLS::Interface and UMLS::Similarity packages

    6. changed the assignSenses function in the AllWords.pm module to return
    a reference to an array rather than the array itself.

    7. moved functions.t to targetword.t and added an allwords.t test case.

  Changes from version 0.07 to 0.09
    1. added a timestamp to the default log directory in the converter
    programs and the umls-allwords-senserelate.pl program

    2. added the --target option in the plain2mm-xml.pl converter program

    3. added a README in the converter directory describing in more detail
    the converter programs (with examples)

    4. added more sample formats in the samples/ directory

    5. added a --candidate option to umls-allwords-senserelate. Note that
    this was originally called --senses but I thought this would be clearer

    6. modified the id to include the document, sentence and term id used in
    the output of the umls-allwords-senserelate.pl program

    7. changes umls-senserelate.pl to umls-targetword-senserelate.pl

  Changes from version 0.05 to 0.07
    1. modified the version information in the .pm modules

  Changes from version 0.03 to 0.05
    1. added an AllWords.pm module with a corresponding
    umls-allwords-senserelate.pl program.

    2. added converter programs in the coverters/ directory

  Changes from version 0.01 to 0.03
    1. modified how the CUIs were obtained when using the vector or lesk
    relatedness measures

  version 0.01
    1. initial release