Real Time Streaming Data Sentiment Analysis

Real Time Company Sentiment Analysis and Forecasting - Day by day sentiment(positive/negative/neutral) of the company (or query).

Installation

• Clone this repo to your local machine using

$ git clone https://github.com/Ar9av/streaming_data_nlp_pipeline.git

change the working directory

$ cd streaming_data_nlp_pipeline

Setup / Requirements

• Install the requirements using the following commands

$ pip install -r requirements.txt

• Make sure you have Redis

In case not , Run the following to install and start redis-server

Redis Installation

$ cd
$ wget http://download.redis.io/releases/redis-6.0.6.tar.gz
$ tar xzf redis-6.0.6.tar.gz
$ cd redis-6.0.6
$ make

Redis Server Start

$ src/redis-server

Once the redis server is started

Configuring the Parameter

Cange the parameters in config.yml and change the COMPANY_NAME to the desired query and run the following command.

• Now in different terminal windows run the following:

$ python consumer/consumer.py

$ python time_series_analytics/time_series.py

$ python producer/prod.py

If the above throws error , probably you have different versions of python installed in your system. Replace python with python3

NOTE: Maintain the order of run to get the result instantenously as new results take time.

The windows should look something like this producing the necessary output.

3 tasks :
1) Scraping Data (prod.py)
2) Sentiment Analysis (consumer.py)
3) Time Series Forecasting (time_series.py)

Run in async matter but waits for the response from others through 3 redis channels which is used for publishing and listening.