Voice based sentiment analysis using python

 Hello python programmers this is AK, In this blog, I'm going to show how to 

perform a voice-based sentiment analysis using python. 

In this project, we’re going to use two libraries in python. 

The first one is the Vader sentiment package and the second one is the

speech recognition library in python.

 

 

Both two are open source libraries; you can use these packages an unlimited 

number of times.

Sentiment analysis (or opinion mining) is a natural language processing technique 

used to determine whether data is positive, negative, or neutral. Here the data means

 your comments about any particular product or a review kind of thing.

For example, Take a company like Amazon that sells so many products to people. 

And if you want to buy any product from amazon the first thing you will search for

 the reviews about the product. The reviews are very important 

and they will give an impression about the 

product to the buyer.

Companies like Amazon are using the sentiment analysis method to give

 the product rating. It will be very helpful for the buyers 

to understand more about the product.

 

The Internet is free for everyone and you can type anything about any particular 

product on e-commerce sites. Customers express their thoughts and feelings 

more openly than ever before, sentiment analysis is becoming an essential tool to 

monitor and understand those opinions about the product. Automatically analyzing

 customer feedback, such as opinions in survey responses, social media conversations, 

and product ratings, allows brands to learn what makes customers happy or frustrated

 so,that they can tune their products and services to meet their customers’ needs.

So this is what the sentiment analysis is

 

Why sentiment analysis is very important for business?

 

A company must release the customers' needed product and that is the sign of

 good companies. The customers are the first priority to the MNC companies.

These companies are tracking the customers data. So Sentiment  

analysis is extremely important for that cause  because it helps to quickly understand 

the overall opinions of their customers. If you understand your customers' needs

 you can sell those products to your customers and it will be very helpful to grow your

 business .So this is the simple concept about sentimental analysis.

Let’s start to code.

Basically sentiment analysis is a simple project in the field of Machine learning.

But this blog is not a machine learning subject here im using python library to do 

this job.Sentiment analysis is a simple project in python.Suppose if you want the

 machine learning concepts by applying in sentiment analysis comment down your

 thoughts 

So i will make this sentiment analysis project by applying the machine learning 

concepts in python. 

 

STEP 1:

 

pip install vadersentiment

pip install speech recognition 
 

First you need to install a python package called Vader sentiment.So open your

 command prompt and just type pip install vadersentiment.So this will help you to

 install the library.

And you’ve to install the speech-recognition library so type

 pip install speech-recognition this will install this library.

After installing the packages you need to import the vadersentiment library and 

speech-recognition library on your IDE.

 

STEP 2:

 

from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
import speech_recognition as sr 

 

And create a variable called a recognizer and call the instance recognizer() this will 

help to initiate the speech-recognition library.

 

STEP 3:

recognizer=sr.Recognizer()
with sr.Microphone() as source: 

   print('Clearing background noise...')
recognizer.adjust_for_ambient_noise(source,duration=1)
print('Waiting for your message...')
recordedaudio=recognizer.listen(source)
print('Done recording..') 

Next we want to initiate our microphone for recording our voices.For that type

 with sr.Microphone as source.Here we’re declaring 

the instance microphone as source variable.And In this block 

of code you’ve to clear your background noises.

In this library there is an option called adjust for ambient_noises.This is an object

 and here just pass the parameter as your source.This object simply clears the

 background noises before starting to record.

Now we’ve to record the voices for that

 create variable recorded 

audio and in this variable you need to pass the recognizer.listen object 

and inside this pass your source.

These two statements are very important when doing the speech-recognition 

related tasks on your project. Mainly these statements 

help to increase the quality

 of recording your voices.

 

STEP 4: 

 

try:
print('Printing the message..')
text=recognizer.recognize_google(recordedaudio,language='en-US')
print('Your message:{}'.format(text))
except Exception as ex:
print(ex)

 

We finished the recording process in our code.Now you need to create a try block 

and here create variable text and call recognize_google object.There are so many

 APIs are available in this library.The recognize_google is a most used voice_recognizer 

instance in this library.So I'm using this one to recognize my voices.And I created the

 except block this will print any errors or exception in our code.

 

Now we finished the speech_recognition tasks.Let’s write the code for the 

sentiment analysis.

 

STEP 5:

#Sentiment analysis

Sentence=[str(text)]
analyser=SentimentIntensityAnalyzer()
for i in Sentence:
v=analyser.polarity_scores(i)
print(v) 

Declare a variable called sentence and make sure this variable must be a list datatype

.If you’re passing the stand-alone string it will take each character and do the analysis 

with each character  that is present on the string.

It’s not useful one when we do this.Create a list in sentence variable and pass

 the str(text) this is the correct form.

And create a variable analyser and call the instance sentimentInstentsityAnalyzer().

Next we’ve to traverse the list.If you want to perform the sentiment analysis on 

single sentence you don’t need to use the list and  forloop condition.

If you want to do the sentiment analysis for a group of sentences then this will be

 the optimal way to analyze the sentence.And create a variable V and in this variable 

you need to call the analyzer.polarity_scores.The polarity scores are very important to

  determine whether the sentence is positive or negative or neutral one.And in this 

object pass the sentence variable.

So that’s all about this code, let's run this program.

 

 

When you run this program you will get the output like positive, negative or neutral.

Based on these category values you can easily predict whether you are telling a 

positive comment or negative comment or neutral comment.

 

 

Thanks for reading this article💗

 

 BUY ME A CUP OF COFFEE😻

 

Full code :Voice based sentiment Analysis