Apache 临床的文本分析和知识提取系统 (cTAKES) 是一个开源的自然语言处理系统，用于从临床电子病历自由文本中抽取信息。可处理临床笔记抽取出统一的医学语言系统(UMLS)——药物,疾病/障碍,标志/症状,解剖网站和程序等。

Apache cTAKES 3.1.1 改进内容包括：

Sub-task

• [CTAKES-150] - Apostrophe parsing is not thread safe because of global variable

Bug

• [CTAKES-242] - XCasWriterCasConsumer issues with Apache cTAKES 3.1
• [CTAKES-244] - error using relation extractor launch (see also CTAKES-72)
• [CTAKES-245] - template filler - ClassNotFoundException: EntityMentionPairRelationExtractorAnnotator
• [CTAKES-251] - Infinite recursion in DrugMentionAnnotator.java
• [CTAKES-253] - YTEX ctakes patches port
• [CTAKES-254] - Apostrophe in contraction breaks TokenizerPTB
• [CTAKES-266] - tokenizer creates empty tokens before contractions
• [CTAKES-270] - ctakes-assertion-zoner missing from trunk's bin.xml in ctakes-distribution
• [CTAKES-271] - build fails due to mismatches within expectedoutput directory - e.g Sentence model was changed in 3.1.0
• [CTAKES-272] - regression test run during build of 3.1.1 fails with NoClassDefFoundError: CharacterOffsetToLineTokenConverter

Improvement

• [CTAKES-123] - README should mention how to get LVG database if it is going to mention how to get UMLS dictionaries
• [CTAKES-219] - configure log4j for ctakes-assertion module and remove stdout print statements
• [CTAKES-243] - Delete multiple copies of SentenceDetectorAnnotator.xml
• [CTAKES-264] - chunk adjuster iterates inefficiently
• [CTAKES-269] - log4j.xml filenotfound

New Feature

• [CTAKES-257] - utilities to turn dependency graphs into SimpleTree

Task

• [CTAKES-256] - Add a test doc that contains various edge cases for regression testing
• [CTAKES-259] - create a Maven profile for nightly builds that excludes the distribution module (.zips)
• [CTAKES-267] - create a 3.1.1 patch release