Worth taking a look at the unicode sentence segmentation algorithm rules: http://unicode.org/reports/tr29/#Sentence_Boundaries http://unicode.org/reports/tr29/#

Worth taking a look at the unicode sentence segmentation algorithm rules: http://unicode.org/reports/tr29/#Sentence_Boundaries http://unicode.org/reports/tr29/#Sentence_Boundaries Also at the CLDR sentence break supressions: http://unicode.org/cldr/trac/browser/tags/release-27-0-1/common/segments/en.xml http://unicode.org/cldr/trac/browser/tags/release-27-0-1/com... If your rules treat an edge case that the above don't it'd probably be worth trying to suggest improvements to the unicode rules or the locale-specific ones.