Stop Words Cleaner for Persian

Description

This model removes ‘stop words’ from text. Stop words are words so common that they can be removed without significantly altering the meaning of a text. Removing stop words is useful when one wants to deal with only the most semantically important words in a text, and ignore words that are rarely semantically relevant, such as articles and prepositions.

Live Demo Open in Colab Download Copy S3 URI

How to use

...
stop_words = StopWordsCleaner.pretrained("stopwords_fa", "fa") \
        .setInputCols(["token"]) \
        .setOutputCol("cleanTokens")
nlp_pipeline = Pipeline(stages=[document_assembler, tokenizer, stop_words])
light_pipeline = LightPipeline(nlp_pipeline.fit(spark.createDataFrame([['']]).toDF("text")))
results = light_pipeline.fullAnnotate("جان اسنو غیر از سلطان شمال ، یک پزشک انگلیسی و رهبر توسعه بیهوشی و بهداشت پزشکی است.")

...
val stopWords = StopWordsCleaner.pretrained("stopwords_fa", "fa")
        .setInputCols(Array("token"))
        .setOutputCol("cleanTokens")
val pipeline = new Pipeline().setStages(Array(document_assembler, tokenizer, stopWords))
val data = Seq("جان اسنو غیر از سلطان شمال ، یک پزشک انگلیسی و رهبر توسعه بیهوشی و بهداشت پزشکی است.").toDF("text")
val result = pipeline.fit(data).transform(data)

import nlu

text = ["""جان اسنو غیر از سلطان شمال ، یک پزشک انگلیسی و رهبر توسعه بیهوشی و بهداشت پزشکی است."""]
stopword_df = nlu.load('fa.stopwords').predict(text)
stopword_df[["cleanTokens"]]

Results

[Row(annotatorType='token', begin=0, end=2, result='جان', metadata={'sentence': '0'}),
Row(annotatorType='token', begin=4, end=7, result='اسنو', metadata={'sentence': '0'}),
Row(annotatorType='token', begin=9, end=11, result='غیر', metadata={'sentence': '0'}),
Row(annotatorType='token', begin=16, end=20, result='سلطان', metadata={'sentence': '0'}),
Row(annotatorType='token', begin=22, end=25, result='شمال', metadata={'sentence': '0'}),
...]

Model Information

Data Source

The model is imported from https://github.com/WorldBrain/remove-stopwords