Sentiment Analysis

This application analysis texts to detect if they have negative or positive sens.

Comment

Inspiration

More and more, the economic, social, political and health situation is deteriorating in the world. As a result, we are seeing an exponential rise in violence (ciber violence) on social networks, especially on social medias. This is what inspired me to work on such a project

What it does

It analysis the sentiment contained in texts. First , You have a form where you can write a text or copy and past a text an then hit the Submit button. Then the text is analysed and the result is displayed showing you whether the content of the text is positive, negative to neutral. Secondly , you can upload a dataset (excel file, csv file) of multiple texts (in our test for example, we used a dataset (tweets) from twitter's posts. Then the data content in it is analysed, each tweet is analysed and you get the result. You can display the negative content, the positive content, the users who wrote them, the place from where they wrote it and the date.

How we built it

I built it with python and Streamlit framework. I first installed the python virtual environment, I created a virtual machine then I activated it, Then I installed the streamlit framework and created the project I installed the relevant libraries (to be seen in the requirement.txt file) I created the different components (img, templates...) I implemented the form with the textarea field I implemented the sentiment analysis using the text provided in the form textarea I implemented the component allowing the user to upload a file from the computer I finally implemented the deep learning model to analyse the whole data and display the data statistics

Challenges we ran into

The main challenge was to well implement the model. Because, some textts can content positive and negative expressions, so I had to deal with such situations to determine whether the global sense of the sentence is positive or negative.

Accomplishments that we're proud of

I'm proud to have implemented a model that can literally simulate human's brain

What we learned

I learned a lot about the_** vaderSentiment**_ librairy.

What's next for Sentiment Analysis

Next, I will ask people to test and have feedback; If they're satisfied about it, I'll add some changes and put it in production so that everyone interested can use it.

Built With

  • awesome-streamlit
  • csv
  • excel
  • logging
  • numpy
  • pandas
  • pil
  • python
  • resources
  • sentimentintensityanalyzer
  • streamlit
  • vadersentiment
  • virtualenv
Share this project:

Updates