http://pangusegment.codeplex.com
PanGu.dll 调用方法
初始化
在进程启动时,我们需要对盘古分词进行初始化,初始化的调用代码如下:
默认方式初始化
PanGu.Segment.Init();
这个调用将使用和 PanGu.dll 同路径下的 pangu.xml 配置文件
指定配置文件方式初始化
PanGu.Segment.Init(filename);
filename 为pangu.xml 的完整路径名,如“c:pangu.xml”
在某些应用中,pangu.xml 不一定在pangu.dll相同的路径下,或者无法获取pangu.dll的当前路径,这时需要通过这种方式调用来让调用者指定盘古分词所用的配置文件的绝对路径。
分词
Segment segment = new Segment(); ICollection<WordInfo> words = segment.DoSegment(text);
或
ICollection<WordInfo> words = segment.DoSegment(text, options);
或
ICollection<WordInfo> words = segment.DoSegment(text, options, parameters);
其中
· text 为需要分词的文本
· options 为自定义分词选项,默认为pangu.xml 中指定的分词选项
· parameters 为分词参数,默认为pangu.xml 中指定的分词参数
分词选项定义:
public class MatchOptions { /// <summary> /// 中文人名识别 /// </summary> public bool ChineseNameIdentify = false; /// <summary> /// 词频优先 /// </summary> public bool FrequencyFirst = false; /// <summary> /// 多元分词 /// </summary> public bool MultiDimensionality = true; /// <summary> /// 英文多元分词,这个开关,会将英文中的字母和数字分开。 /// </summary> public bool EnglishMultiDimensionality = false; /// <summary> /// 过滤停用词 /// </summary> public bool FilterStopWords = true; /// <summary> /// 忽略空格、回车、Tab /// </summary> public bool IgnoreSpace = true; /// <summary> /// 强制一元分词 /// </summary> public bool ForceSingleWord = false; /// <summary> /// 繁体中文开关 /// </summary> public bool TraditionalChineseEnabled = false; /// <summary> /// 同时输出简体和繁体 /// </summary> public bool OutputSimplifiedTraditional = false; /// <summary> /// 未登录词识别 /// </summary> public bool UnknownWordIdentify = true; /// <summary> /// 过滤英文,这个选项只有在过滤停用词选项生效时才有效 /// </summary> public bool FilterEnglish = false; /// <summary> /// 过滤数字,这个选项只有在过滤停用词选项生效时才有效 /// </summary> public bool FilterNumeric = false; /// <summary> /// 忽略英文大小写 /// </summary> public bool IgnoreCapital = false; /// <summary> /// 英文分词 /// </summary> public bool EnglishSegment = false; /// <summary> /// 同义词输出 /// </summary> /// <remarks> /// 同义词输出功能一般用于对搜索字符串的分词,不建议在索引时使用 /// </remarks> public bool SynonymOutput = false; /// <summary> /// 通配符匹配输出 /// </summary> /// <remarks> /// 同义词输出功能一般用于对搜索字符串的分词,不建议在索引时使用 /// </remarks> public bool WildcardOutput = false; /// <summary> /// 对通配符匹配的结果分词 /// </summary> public bool WildcardSegment = false; /// <summary> /// 是否进行用户自定义规则匹配 /// </summary> public bool CustomRule = false; }
分词参数定义
[Serializable] public class MatchParameter { /// <summary> /// 多元分词冗余度 /// </summary> public int Redundancy = 0; /// <summary> /// 未登录词权值 /// </summary> public int UnknowRank = 1; /// <summary> /// 最匹配词权值 /// </summary> public int BestRank = 5; /// <summary> /// 次匹配词权值 /// </summary> public int SecRank = 3; /// <summary> /// 再次匹配词权值 /// </summary> public int ThirdRank = 2; /// <summary> /// 强行输出的单字的权值 /// </summary> public int SingleRank = 1; /// <summary> /// 数字的权值 /// </summary> public int NumericRank = 1; /// <summary> /// 英文词汇权值 /// </summary> public int EnglishRank = 5; /// <summary> /// 符号的权值 /// </summary> public int SymbolRank = 1; /// <summary> /// 强制同时输出简繁汉字时,非原来文本的汉字输出权值。 /// 比如原来文本是简体,这里就是输出的繁体字的权值,反之亦然。 /// </summary> public int SimplifiedTraditionalRank = 1; /// <summary> /// 同义词权值 /// </summary> public int SynonymRank = 1; /// <summary> /// 通配符匹配结果的权值 /// </summary> public int WildcardRank = 1; /// <summary> /// 过滤英文选项生效时,过滤大于这个长度的英文。 /// </summary> public int FilterEnglishLength = 0; /// <summary> /// 过滤数字选项生效时,过滤大于这个长度的数字。 /// </summary> public int FilterNumericLength = 0; /// <summary> /// 用户自定义规则的配件文件名 /// </summary> public string CustomRuleAssemblyFileName = ""; /// <summary> /// 用户自定义规则的类的完整名,即带名字空间的名称 /// </summary> public string CustomRuleFullClassName = ""; }
返回为WordInfo 的集合
public class WordInfo : WordAttribute, IComparable<WordInfo> { /// <summary> /// 当前单词类型 /// </summary> public WordType WordType; /// <summary> /// 原始的单词类型 /// </summary> public WordType OriginalWordType; /// <summary> /// 单词在text 中的起始位置 /// </summary> public int Position; /// <summary> /// Rank for this word /// 单词权重 /// </summary> public int Rank; /// <summary> /// 单词 /// </summary> public String Word; /// <summary> /// 词性 /// </summary> public POS Pos; /// <summary> /// 词频 /// </summary> public double Frequency; }
配置文件PanGu.xml
<?xml version="1.0" encoding="utf-8"?> <PanGuSettings xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns="http://www.codeplex.com/pangusegment"> <DictionaryPath>..Dictionaries</DictionaryPath> <MatchOptions> <ChineseNameIdentify>true</ChineseNameIdentify> <FrequencyFirst>false</FrequencyFirst> <MultiDimensionality>false</MultiDimensionality> <FilterStopWords>true</FilterStopWords> <IgnoreSpace>true</IgnoreSpace> <ForceSingleWord>false</ForceSingleWord> <TraditionalChineseEnabled>false</TraditionalChineseEnabled> <OutputSimplifiedTraditional>false</OutputSimplifiedTraditional> <UnknownWordIdentify>true</UnknownWordIdentify> <FilterEnglish>false</FilterEnglish> <FilterNumeric>false</FilterNumeric> <IgnoreCapital>false</IgnoreCapital> <EnglishSegment>false</EnglishSegment> <SynonymOutput>false</SynonymOutput> <WildcardOutput>false</WildcardOutput> <WildcardSegment>false</WildcardSegment> <CustomRule>false</CustomRule> </MatchOptions> <Parameters> <UnknowRank>1</UnknowRank> <BestRank>5</BestRank> <SecRank>3</SecRank> <ThirdRank>2</ThirdRank> <SingleRank>1</SingleRank> <NumericRank>1</NumericRank> <EnglishRank>5</EnglishRank> <EnglishLowerRank>3</EnglishLowerRank> <EnglishStemRank>2</EnglishStemRank> <SymbolRank>1</SymbolRank> <SimplifiedTraditionalRank>1</SimplifiedTraditionalRank> <SynonymRank>1</SynonymRank> <WildcardRank>1</WildcardRank> <FilterEnglishLength>0</FilterEnglishLength> <FilterNumericLength>0</FilterNumericLength> <CustomRuleAssemblyFileName>CustomRuleExample.dll</CustomRuleAssemblyFileName> <CustomRuleFullClassName>CustomRuleExample.PickupNokia</CustomRuleFullClassName> <Redundancy>0</Redundancy> </Parameters> </PanGuSettings>
其中DictionaryPath 指明字典所在目录,可以为相对路径也可以为绝对路径。
MatchOptions 对应分词选项
Parameters 对于分词参数
高亮组件PanGu.HighLight.dll 调用方法
//创建HTMLFormatter,参数为高亮单词的前后缀 PanGu.HighLight.SimpleHTMLFormatter simpleHTMLFormatter = new PanGu.HighLight.SimpleHTMLFormatter("<font color="red">", "</font>"); //创建Highlighter ,输入HTMLFormatter 和盘古分词对象Semgent PanGu.HighLight.Highlighter highlighter = new PanGu.HighLight.Highlighter(simpleHTMLFormatter, new Segment()); //设置每个摘要段的字符数 highlighter.FragmentSize = 50; //获取最匹配的摘要段 String abstract = highlighter.GetBestFragment(keywords, news.Content);