Skip to main content

Python package for exploratory text data analysis

Project description

Arabica

Python package for exploratory text data analysis

Text data is often recorded as a time series with significant variability over time. Some examples of time-series text data include Twitter tweets, product reviews, and newspaper headlines. Arabica provides functions to make the exploratory analysis of such datasets simple.

Arabica provides these methods:

  • arabica_freq: calculates unigram, bigram, and trigram frequencies over a period (year, month, day)

It can apply all or a selected combination of the following cleaning operations:

  • Remove digits from the text
  • Remove punctuation from the text
  • Remove standard list of stopwords
  • Remove an additional specific list of words

Arabica works with texts of languages based on the Latin alphabet, uses clean-text for punctuation cleaning, and enables stop words removal for languages in the NLTK corpus of stopwords.

It reads dates in standard date and datetime formats (e.g., 2013–12–31, 2013/12/31, 09-Feb-2009, 2013–12–31 11:46:17, 09/02/2009 09:26). It is preferable to use the US-style dates (MM/DD/YYYY) rather than the European-style date format (DD/MM/YYYY).

Installation

Arabica requires Python >=3.7, NLTK, clean-text, and numpy to execute. To install using pip, use:

pip install arabica

Usage

  • Import the library:
from arabica import arabica_freq
  • Choose a method:

arabica_freq returns a dataframe with aggregated unigrams, bigrams, and trigrams frequencies over a period. To remove stopwords, select aggregation period, and choose a specific set of cleaning operations:

def arabica_freq(text: str,                # Text
                 time: str,                # Time
                 time_freq: str ='',       # Aggregation period: 'Y'/'M'/'D', if no aggregation: 'ungroup'
                 max_words: int ='',       # Max number for unigrams, bigrams and trigrams displayed
                 stopwords: [],            # Languages for stop words
                 skip: [],                 # Remove additional strings
                 numbers: bool = False     # Remove all digits
                 punct: bool = False,      # Remove all punctuation
                 lower_case: bool = False, # Lowercase text before cleaning and frequency analysis
) 

A list of available languages for stopwords is printed with:

from nltk.corpus import stopwords
print(stopwords.fileids())

It is possible to remove more sets of stopwords at once by stopwords = ['language 1', 'language2','etc..']

Examples

Time-series n-gram analysis

Returns a table with unigram, bigram, and trigram frequencies over a period of time.

import pandas as pd
from arabica import arabica_freq
data = pd.DataFrame({'text': ['The ordering process was very easy & straight forward. They have great customer service and sorted any issues out very quickly.',
                              'So far seems to be the wrong product for me :-/ grrrrr...',
                              'Excellent, service, thank you really, really, really much!!!'],
                     'time': ['2013-08-8', '2013-09-8','2014-10-8']})
arabica_freq(text = data['text'],
             time = data['time'],
             time_freq = 'M',           # Calculates monthly n-gram frequencies
             max_words = 2,             # Displays two most frequent unigrams, bigrams, and trigrams
             stopwords = ['english'],   # Removes English set of stopwords
             skip = ['grrrrr'],         # Excludes string from n-gram calculation
             numbers = True,            # Removes numbers
             punct = True,              # Removes punctuation
             lower_case = True)         # Lowercase text before cleaning and n-gram calculation  

Descriptive n-gram analysis

Returns unigram, bigram, and trigram frequencies without period aggregation.

arabica_freq(text = data['text'],
             time = data['time'],
             time_freq = 'ungroup',        # No aggregation made
             max_words = 2,
             stopwords = ['english'],
             skip = ['grrrrr'],       
             numbers = True,
             punct = True
             lower_case = True)

Documentation and tutorials

Read the documentation here. For more examples of coding, read this tutorial:

Text as Time Series: Arabica 1.0.0 Brings New Features for Exploratory Text Data Analysis here

License

MIT

For any questions, issues, bugs, and suggestions, please visit here.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

arabica-1.0.5.tar.gz (8.8 kB view hashes)

Uploaded Source

Built Distribution

arabica-1.0.5-py3-none-any.whl (7.8 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page