i-am-sayantan/public-sentiment-analysis-based-on-twitter-hashtags

This project focuses on sentiment analysis. Social Sentiment analysis is the use of natural language processing (NLP) to analyze social conversations online and determine deeper context as they apply to a topic, brand or theme.

❄️Public-sentiment-analysis-based-on-twitter-hashtags❄️

Discord server for communication purposes: https://discord.gg/fKdd2tHeSR

ABSTRACT 📄 :

Sentiment analysis in reviews, comments, tweets, captions is one of the trending projects right now in the DL and ML domains, we use the NLU engine to analyze the sentiments of the texts, as well as try to classify the labelled texts using BERT, Roberta models. This project focuses on the unlabeled where you can try to analyze the tweets of the hashtags using different libraries and based on your analysis you can make plots and make an observation from it. One can understand the behaviour of the public sentiment at different times, let's say what is the trend of positivity in human sentiments during weekdays and weekends, how is the trend of the negativity curve during the month ends, etc. For completing the goals datasets have been provided, as well as a demo is also added that how the datasets had been made.

TASKS ✏️ :

1. Use different natural language understanding scoring methods or libraries to determine the sentiment score of the hashtags. Put your codes in a Jupyter Notebook. Here is a video about the NLU and sentiments analysis: Sentiment analysis types and approaches
2. If you are done with part 1 of the task you can make the plots of the scores, based on weeks, days and months, keep them in part 2 of the same notebook. Here is a video, which will help you to create plots: Creating and Customizing Our First Plots
3. If you want to do more after part 2 you can make a small report of your understanding based on your plots observations, write some discussions about them and that will be the end of the project. You can have a look on this video: Sentiment Analysis: extracting emotion through machine learning | Andy Kim | TEDxDeerfield

HOW TO CONTRIBUTE 😃 :

1. First go to , where you can find the issues. Comment on the task 1 issue, if your are interested.
2. You can fork and start working on the project, when you are done make a PR on the ,please dont make the PR in the main branch.
3. For task 1 and task 2 only one jupyter notebook is needed and make you PR inside , the name of the file should be name_of_the_contributor.ipynb, and should be in this format:
4. For task 3 please make your PR inside , the name of the file should be name_of_the_contributor.pdf, and should be in this format:
5. Please don't forget to add a comment on the issues before making the PRs, for questions please feel free to drop it in the discord channel.

ABOUT THE DATASET 📈:

• To use the dataset go inside the and get the understanding from the

• If you wonder how the datasets had been made, have a look inside this and have a look at the example.ipynb, if you wants to test that notebook please run this command first:

  pip install -r requirements.txt
  

  And then open jupyter notebook and test them with your examples, have fun!!!!:wink::wink::wink:

SUPPORTING MATERIALS :

• Links to some research paper: Link 1 : DepecheMood++,a Bilingual Emotion Lexicon Built Through Simple Yet Powerful Techniques ; Link 2 : Multilingual Twitter Sentiment Classification ; Link 3 : Sentiment Analysis in Social Media and Its Application; Link 4 : A Study on Sentiment Analysis Techniques of Twitter.

• Links to the notebooks: Notebook1 ; Notebook2 ; Notebook3 ; Notebook4.

CONTRIBUTORS :

Made with ❤️ by

public-sentiment-analysis-based-on-twitter-hashtags's Issues

Improve the Readme.md and data.md.

📑 Documentation

If you are able to understand the tasks of the project, then you add some more parts in the Readme.md and in data.md

1. as you can add new parts in the readme, like research papers in this domain, kaggle solutions and Github links, and put the name as supporting material
2. You can make the abstract better.
3. Add more points to the data.md and describe more about the columns, as well more observational points.

Make sentiment scores from the tweets of the hashtags using dataset.csv

📝 TASK 1:

Make a jupyter notebook, in which you can write the codes to find the sentiment scores of the tweets from the dataset.csv.

• About the task : For better understanding, please have a look here: TASK 1
• contribution : For making a contribution please have look here: Contribute
• Format : Format of the jupyter notebook should be this : only till part 1, and name should be Name_of_contributer.ipynb
• Output : Here is a sample snippet of the output after adding the scores

Since there are multiple algorithms for making the scores, multiple contributors can participate in it and make their own notebook as a contribution.

make the datasets better!!!

To help the contributors make better versions of datasets and compress them into a RAR file.

Make a small report of the Task 2 plot observations

📝 TASK 3:

After completing #4 you can make a report from the plots and complete task 3 of this project.

• About the task: For better understanding, please have a look here: TASK 3
• contribution : For making a contribution please have a look here: Contribute
• Format: For this task you need to make a report, the format of the report should be this : and name should be Name_of_contributer.pdf.

Since a person can make different observations from plots and solve different unsolved questions, multiple contributors can participate in it and make their own notebook as a contribution.

Make plots from the scores created from TASK 1

📝 TASK 2:

After completing #3 you can make the plots from the scores and complete task 2 of this project. This task should be done in the same notebook in which you had done task 1.

• About the task : For better understanding, please have a look here: TASK 2
• contribution : For making a contribution please have look here: Contribute
• Format : Format of the jupyter notebook should be this : part1 + part 2, and name should be Name_of_contributer.ipynb
• Output: You can make line plots, bar plots, pie charts etc from the dataset.

Since a person can make multiple plots and make different observations from them, multiple contributors can participate in it and make their own notebook as a contribution.