NLG model

Natural Language Generation model which produces a report from a formatted input table whether excel or tableau, and it is used pandas.DataFrame to access the data and produces a summarized report.

Pratical Example

Input	Output

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

During the development of the project it was used python3 (3.8.4)

Installing

Get pip3:

python3 get-pip.py

Get technologies

pip3 install Flask
pip3 install openpyxl
pip3 install excel2img
pip3 install elasticsearch
pip3 install nltk
pip3 install wordnet
pip3 install matplotlib
pip3 install numpy
pip3 install pandas
pip3 install sklearn

Running the offline model

The head file is local_test.py and contains several examples in json format. Due to confidentiality with the company I am not allowed to publish the server version so, in order to run the offline model it is needed to convert the table to json in the following format:

variable name: {'data':[     <table content>    ]}

% EXAMPLE

mammals3 = {
    "data": [
        {
            "Animal": "Okapi",
            "Adult weight(kg)": "225",
            "Maximum longevity(y)": "34",
            "Female maturity(d)": "578",
            "Male maturity(d)": "790",
            "Litter size": "1",
            "Domain": "Earth"
        }
    ]
}

When WTITE_FILE is True, it produces an HTML

Running the tests

Here is more Tableaus which the model produces nice reports:

• https://public.tableau.com/views/money1_13/CashInstruments?:embed=y&:display_count=y&:origin=viz_share_link
• https://public.tableau.com/profile/venkat8761#!/vizhome/worldCovid-19/Covid-19
• https://public.tableau.com/views/RegisteredVehiclesOpenDataProject/BrandBenchmark?:showVizHome=n&:embed=t
• https://public.tableau.com/views/ThePulpFictionConnection/PulpFictionConnection?:showVizHome=n&:embed=t
• https://public.tableau.com/views/CashlessSociety/CashlessSociety?:showVizHome=n&:embed=t
• https://public.tableau.com/en-us/gallery/costs-using-car?tab=viz-of-the-day&type=viz-of-the-day
• https://public.tableau.com/views/Womensrepresentationinpoliticsvizforsocialgood/WomeninPolitics?:showVizHome=n&:embed=t
• https://public.tableau.com/views/TopLinkedInSkillsfor20142015and2016/LinkedInDashboard?:showVizHome=n&:embed=t
• https://public.tableau.com/views/BigBookofLineCharts/BBLC1?:showVizHome=n&:embed=t
• https://public.tableau.com/views/TheMeatMap/meat-dash?:showVizHome=n&:embed=t
• https://public.tableau.com/views/2018W27NewYorkRatSightings/NewYorkRatSightings?:showVizHome=n&:embed=t

Documentation

Here is all the documentation that have been supporting the project throughout the years.

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgment

I worked two months in this NLG model during my Summer Intership at Syone at 2020.