sentifish

sentifish-py

sentifish is a Python library for Sentiment analysis of textual data(only English).By using sentifish it is very easy to perform tasks such as part-of-speech tagging, noun phrase extraction, sentiment analysis, classification etc. A good inforamtion for users is that this new version of sentifish has greater speed of sentiment analysis as compare to older versions.

Installation

sentifish-py can be installed using pip similar to other Python packages. Do not use sudo with pip.

To install sentifish-py, simply:

.. code-block:: bash

$ pip install sentifish

Getting Started

To ensure you have installed sentifish successfully you can run the following command in the Python IDLE.

.. code-block:: python

>>> import sentifish

Sentifish have some methods classes which are following described.

sentTokenizer(paragraph)

sentTokenizer( ) is a method. It takes a paragraph as input and then returns a list of sentences of input paragraph.

.. code-block:: python

>>> from sentifish import sentTokenizer
>>> para="This is the first sentence. This is the second sentence.
this is the third sentence."
>>> para_lines=sentTokenizer(para)
>>> para_lines
['This is the first sentence.', 'This is the second sentence.', 'this is the third sentence.']

wordTokenizer(sentence)

wordTokenizer( ) is a method. It takes a paragraph or sentence as input and then returns a list of words, symbols, and numbers of input paragraph or sentence.

.. code-block:: python

>>> from sentifish import wordTokenizer
>>> sent="This is an example sentence."
>>> word_list=wordTokenizer(sent)
>>> word_list
['This', 'is', 'an', 'example', 'sentence', '.']

Class Sentiment( )

Sentiment( ) is a class. By using this class we can find the sentiment of a texual data(it may be a word, sentence or a paragraph). This class has a constructor init(self,text) which takes the text data at the time of instantiation of Sentiment( )

.. code-block:: python

>>> from sentifish import Sentiment
>>> text="Ram is a good boy and he always remains happy"
>>> obj=Sentiment(text)

analyze( )

Class Sentiment( ) has a method analyze( ) it returns a float number in between -1 to +1. +1 for strongly positive sentiment, 0 for neutral and -1 for strongly negative sentiment.

.. code-block:: python

>>> from sentifish import Sentiment
>>> text="Ram is a good boy and he always remains happy"
>>> obj=Sentiment(text)
>>> polarity = obj.analyze( )
>>> polarity
0.75

isPositive( )

Class Sentiment( ) has a method isPositive( ) it return True if the sentiment of the input text is positive. Otherwise it returns False.

.. code-block:: python

>>> from sentifish import Sentiment
>>> text="Ram is a good boy and he always remains happy"
>>> obj=Sentiment(text)
>>> polarity = obj.analyze( )
>>> polarity
0.75
>>> obj.isPositive( )
True

isNegative( )

Class Sentiment( ) has a method isNegative( ) it return True if the sentiment of the input text is negative. Otherwise it returns False.

.. code-block:: python

>>> from sentifish import Sentiment
>>> text="Ram is a good boy and he always remains happy"
>>> obj=Sentiment(text)
>>> polarity = obj.analyze( )
>>> polarity
0.75
>>> obj.isPositive( )
True 
>>> obj.isNegative( )
False

isNeutral( )

Class Sentiment( ) has a method isNeutral( ) it return True if the sentiment of the input text is neutral. Otherwise it returns False.

.. code-block:: python

>>> from sentifish import Sentiment
>>> text="Ram is a good boy and he always remains happy"
>>> obj=Sentiment(text)
>>> polarity = obj.analyze( )
>>> polarity
0.75
>>> obj.isPositive( )
True 
>>> obj.isNegative( )
False 
>>> obj.isNeutral( )
False

NOTE:- It can analys text of only english language.

Class PosTag( )

PosTag( ) is a class used for tagging word with part of speech tags. Class PosTag( ) has a constructor which requires list of words at the time of instantiation. The tagged words will store in a list "tagged_words" which can be access by using object of PosTag( ) class.

.. code-block:: python

>>> from sentifish import wordTokenizer
>>> sent="This is an example sentence."
>>> word_list=wordTokenizer(sent)
>>> from sentifish import PosTag
>>> obj=PosTag(word_list)
>>> obj.tagged_words
[('This', 'This', ['NN']), ('is', 'is', ['HV']), ('an', 'an',['IA']),
('example', 'example', ['NN']), ('sentence', 'sentence', ['VB']),('.', '.', ['SYM'])]

Class Characters( )

Characters( ) is a class which has collection of special characters, small alphabets, capital alphabets and detailed information of "pos tags". To find tags use tags( ) method.

.. code-block:: python

>>> from sentifish import Characters
>>> obj = Characters( )
>>> obj.tags( )
{'HV': 'Helping verb', 'WP': 'Wh-Pronoun', 'CD': 'Cardinal number','PR': 'Pronoun', 
'IN': 'Preposition','INV': 'Negative word','INC':'Word enhancing sense of another word', 
'CC': 'Conjunction', 'SYM': 'Symbol','VB': 'Verb base form', 'VBD': 'Verb past form', 
'VBN': 'Verb past participle form', 
'VBZ': 'Verb s/es/ies/ form','VBG': 'Verb ing form', 'JJ': 'Adjective', 'RB': 'Adverb', 
'Nn': 'Noun', 'V': 'Verb', 'NN': 'Noun', 'IA': 'Indefinite articles'}

To find special chars use specialChars( ) method.

.. code-block:: python

>>> obj.specialChars( )
['`', '~', '@', '#', '$', '%', '^', '&', '*', '-', '_', ';', ':',  
'\\', '|', '/', ',', '<', '.', '>', '?', "'", '"', '!', '+', ' ']

Use capitalAlpha( ) and smallAlpha( ) method to get list of capital alphabets and small alphabets respectively.

.. code-block:: python

>>> obj.capitalAlpha( )
['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M',  
'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z']
>>> obj.smallAlpha( )
['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm',  
'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z']

Class FreqDist( )

FreqDist( ) is a class and by using this class we can find the number of occurrence of word, symbol, and number in a sentence. FreqDist( ) class has a constructor which takes a sentences containing words, symbols or numbers and makes a dictionary in which it takes words as key and number of occurrence of word as value.

.. code-block:: python

>>> from sentifish import FreqDist
>>> para="This is the first sentence. This is the second sentence. this is the third sentence."
>>> obj = FreqDist(para)

to obtain dictionary of words or tokens use class variable "words_dict".

.. code-block:: python

>>> from sentifish import FreqDist
>>> para="This is the first sentence. This is the second sentence. 
this is the third sentence."
>>> obj = FreqDist(para)
>>> obj.words_dict
{'This': 2, 'is': 3, 'the': 3, 'first': 1, 'sentence': 3, '.': 3, 'second': 1, 'this': 1, 'third': 1}

To obtain number of distinct words or tokens in the sentence or input text use "dict_size" class variable.

.. code-block:: python

>>> from sentifish import FreqDist
>>> para="This is the first sentence. This is the second sentence. this is the third sentence."
>>> obj = FreqDist(para)
>>> obj.words_dict
{'This': 2, 'is': 3, 'the': 3, 'first': 1, 'sentence': 3, '.': 3, 'second': 1, 'this': 1, 'third': 1}
>>> obj.dict_size
9

most_common(num)

most_common( ) is method which takes an integer number as input and returns a list of tuple of words or tokens which have high frequency in the sentence with their frequency. Number of words in the list will equal to the input integer.

.. code-block:: python

>>> from sentifish import FreqDist
>>> para="This is the first sentence. This is the second sentence. 
this is the third sentence."
>>> obj = FreqDist(para)
>>> obj.words_dict
{'This': 2, 'is': 3, 'the': 3, 'first': 1, 'sentence': 3, '.': 3,'second': 1, 'this': 1, 'third': 1}
>>> obj.dict_size
9
>>> obj.most_common(2)
[('is', 3), ('the', 3)]

least_common(num)

least_common( ) is method which takes an integer number as input and returns a list of tuple of words or tokens which are least common in the sentence. Number of words in the list will equal to the input integer.

.. code-block:: python

>>> from sentifish import FreqDist
>>> para="This is the first sentence. This is the second sentence. this is the third sentence."
>>> obj = FreqDist(para)
>>> obj.words_dict
{'This': 2, 'is': 3, 'the': 3, 'first': 1, 'sentence': 3, '.': 3,'second': 1, 'this': 1, 'third': 1}
>>> obj.dict_size
9
>>> obj.most_common(2)
[('is', 3), ('the', 3)]
>>> obj.least_common(3)
[('second', 1), ('this', 1), ('third', 1)]

Class Lemmatizer( )

Lemmatizer( ) is a class. By using lemmatizer class user can find the base form of verb from any other form of verb.

lemmatize(word)

Lemmatizer( ) class have a method of name lemmatize. It takes a word of other form and returns the base form of word.

.. code-block:: python

>>> from sentifish import Lemmatizer
>>> obj = Lemmatizer( )
>>> obj.lemmatize("went")
‘go’

Class Polarity( )

Polarity( ) is a class. This class is very useful to fix the polarity of words.

fix_polarity(tagged_words_list)

it is an method of the polarity class. It takes tagged words list as input and then fix the sentiment polarity of words and then returns a list.

.. code-block:: python

>>> from sentifish import wordTokenizer
>>> text="Ram is a good boy and he always remains happy"
>>> word_list=wordTokenizer(text)
>>> word_list
['Ram', 'is', 'a', 'good', 'boy', 'and', 'he', 'always', 'remains', 'happy']
>>> from sentifish import PosTag
>>> obj1 = PosTag(word_list)
>>> obj1.tagged_words
[('Ram', 'Ram', ['NN']), ('is', 'is', ['HV']), ('a', 'a', ['IA']), ('good', 'good', ['JJ']), 
('boy', 'boy', ['Nn']), ('and', 'and', ['CC']), ('he', 'he', ['PR']), ('always', 'always', ['RB']), 
('remains', 'remain', ['VBZ']),('happy', 'happy', ['JJ'])]
>>> from sentifish import Polarity
>>> obj.fix_polarity(obj1.tagged_words)
[('Ram', 'Ram', ['NN', 0.0]), ('is', 'is', ['HV', 0.0]), ('a', 'a', ['IA', 0.0]), ('good', 'good', ['JJ', 0.7]),
('boy', 'boy', ['Nn', 0.0]), ('and', 'and', ['CC', 0.0]), ('he', 'he', ['PR', 0.0]),
('always', 'always', ['RB', 0.0]), ('remains', 'remain', ['VBZ',0.0]), ('happy', 'happy', ['JJ', 0.8])]

remove_stopwords(text)

remove_stopwords(text) is a method and it takes text as input and return a list of words after removing stop words from the input text. Stop words are words which have not any sentiment polarity.

.. code-block:: python

>>> from sentifish import remove_stopwords
>>> text="Ram is a good boy and he always remains happy"
>>> remove_stopwords(text)
['Ram', 'good', 'boy', 'remains', 'happy']

A list of stop word can be found from the Words() class.

.. code-block:: python

>>> from sentifish import Words
>>> obj = Words()
>>> obj.stop_words()
['i', 'me', 'my', 'myself', 'we', 'our', 'ours',…………… 'aren']

remove_bitmap(words_list)

remove_bitmap( ) is a method and it takes text as input and return a list of words after removing words of other languages than english.

.. code-block:: python

>>> from sentifish import remove_bitmap