scala-TFIDF（文章关键词提取）

description

Keywords extraction using TFIDF implemented in Scala, supporting both Chinese and English.

中英文关键词提取，使用TFIDF算法。

example

val content = """《偶然》
                  徐志摩
                  我是天空里的一片云，偶尔投影在你的波心,
                  你不必压抑，
                  更无须欢喜，
                  在转瞬间消灭了踪影。
                  你我相逢在黑夜的海上，
                  你有你的，我有我的方向；
                  你记得也好，
                  最好你忘掉，
                  在这交会时互放的光亮！
                  """
TFIDF.getKeywords(content, 2).foreach(println)
/*
    (交会,0.4707520694552998)
    (光亮,0.4707520694552998)
*/

install

You could download it from the release link directly

usage

scala REPL

Load jar package in advance:

lib -> scala -classpath "*.jar"
Welcome to Scala version 2.11.6
scala> import com.wanghuanming.tfidf._
scala> 

Or load jar package in the Scala Shell:

lib -> scala 
Welcome to Scala version 2.11.6
scala> :require ansj_seg-2.0.8.jar
scala> :require nlp-lang-0.3.jar
scala> :require tfidf_2.10-0.0.2.jar
scala> import com.wanghuanming.tfidf._

Notice that, some early Scala version use ":cp" or ":jar" to load jar, instead of ":require".

scala script

lib ➜ cat test.scala
import com.wanghuanming.tfidf._

val content = "我这个程序不能运行？"
val keywords = TFIDF.getKeywords(content, 2)
keywords.foreach(println)
lib ➜ scala -classpath "*.jar" test.scala

sbt

Put the jar package under the lib/ directory.

Something important!

I have computed the IDF for you in advance, but which may not fit for your own requirements.

So, you'd better compute the IDF based on your own corpus like this. Notice that, the parameter of constructCorpus is a directory containing a huge number of documents.

TFIDF.constructIDF("/your/own/corpus/")

The above command will produce a file called "idf.cache", which is the serialized data of IDF.

dependencies

ansj_seg