C#中使用jieba.NET、WordCloudSharp制作词云图

{

varextractor = newTfidfExtractor;

varwordWeight = extractor.ExtractTagsWithWeight(text, 50);

StringBuilder sbr = newStringBuilder;

sbr.Append( "词语");

sbr.Append( ",");

sbr.Append( "权重");

sbr.AppendLine( ",");

foreach( varitem inwordWeight)

{

sbr.Append(item.Word);

sbr.Append( ",");

sbr.Append(item.Weight);

sbr.AppendLine( ",");

}

stringfilename = "关键词权重统计.csv";

File.WriteAllText(filename, sbr.ToString, Encoding.UTF8);

Console.WriteLine( "关键词提取完成："+ filename);

returnwordWeight.ToArray;

}

使用 JiebaNet.Segmenter.Common下的 Counter类统计词频，其实现来自Python标准库的Counter类（具体接口和实现细节略有不同），代码如下：

///<summary>

///分词并统计词频：默认为精确模式，同时也使用HMM模型

///</summary>

///<param name="text"></param>

///<param name="wordWeightAry"></param>

///<returns></returns>

staticKeyValuePair< string, int>[] Counter( stringtext, WordWeightPair[] wordWeightAry)

{

varsegmenter = newJiebaSegmenter;

varsegments = segmenter.Cut(text);

varfreqs = newCounter< string>(segments);

KeyValuePair< string, int>[] countAry = newKeyValuePair< string, int>[wordWeightAry.Length];

for( inti = 0; i < wordWeightAry.Length; i++)

{

stringkey = wordWeightAry[i].Word;

countAry[i] = newKeyValuePair< string, int>(key, freqs[key]);

}

StringBuilder sbr = newStringBuilder;

sbr.Append( "词语");

sbr.Append( ",");

sbr.Append( "词频");

sbr.AppendLine( ",");

foreach( varpair incountAry)

{

sbr.Append(pair.Key);

sbr.Append( ",");

sbr.Append(pair.Value);

sbr.AppendLine( ",");

}

stringfilename = "词频统计结果.csv";

File.WriteAllText(filename, sbr.ToString, Encoding.UTF8);

Console.WriteLine( "词频统计完成："+ filename);

returncountAry;

}

使用 WordCloudSharp生成词云图，蒙版图必须使用黑白图片，记得手动引用 System.Drawing，代码如下：

///<summary>

///创建词云图

///</summary>

///<param name="countAry"></param>

staticvoidCreateWordCloud(KeyValuePair< string, int>[] countAry)

{

stringmarkPath = "mask.jpg";

stringresultPath = "result.jpg";

Console.WriteLine( "开始生成图片，读取蒙版："+ markPath);

Image mask = Image.FromFile(markPath);

//使用蒙版图片

varwordCloud = newWordCloud(mask.Width, mask.Height, mask: mask, allowVerical: true, fontname: "YouYuan");

//不使用蒙版图片

//var wordCloud = new WordCloud(1000, 1000,false, null,-1,1,null, false);

varresult = wordCloud.Draw(countAry.Select(it => it.Key).ToList, countAry.Select(it => it.Value).ToList);

result.Save(resultPath);

Console.WriteLine( "图片生成完成，保存图片："+ resultPath);

}

运行测试

以本文为分析文本生成词云图，代码如下：

staticvoidMain( string[] args)

{

stringtext = File.ReadAllText( "待处理数据.txt");

varwordWeight = ExtractTagsWithWeight(text);

varwordFreqs = Counter(text, wordWeight);

CreateWordCloud(wordFreqs);

Console.Read;

}

蒙版图如下：

词云图如下（使用蒙版）：

词云图如下（不使用蒙版）：

在得到关键词的词频信息后，通 过在线工具网站生成词云图片会更加方便一点，如词云文字、图悦等。

参考资料 返回搜狐，查看更多

• jieba.NET是jieba中文分词的.NET版本（C#实现）
• TF-IDF算法和TextRank算法的分析比较
• Python生成词云图