Determining the incentive of chats

[mohitmishra786]

• We can export the chat and then try to check whether the chats are hate-related or non-hate.

[JRavi2]

Are you suggesting we perform some kind of sentiment analysis on the messages? Like figuring out how many negative/positive messages are there and from which users? That could be an interesting feature!

[mohitmishra786]

Yes

[JRavi2]

Great! Would you be interested to work on it?

[mohitmishra786]

@JRavi2 Yes

I have applied for this using KWOC, so how to proceed with it

[JRavi2]

@mohitmishra786 Awesome!

To get started, I'd recommend you to try running the program, and then try to understand the code by understanding the flow of data. The main entry point and all the functionality related to parsing/importing of data is present in the src/chat_analyzer.py file, whereas all the data analysis part lies in the src/chat_funtions.py file.

Once you have a decent understanding of the code, you could start implementing the feature. Feel free to ping me anytime if you get stuck somewhere :)

[asaraf7]

Can only those people work on an issue who have been assigned?

[JRavi2]

@asaraf7 Yeah, @mohitmishra786 is working on this. I see that you've shown interest in two other issues, so let's get those sorted out first :)

[asaraf7]

Hey, I can perform sentiment analysis, but how will I extract the messages, given that format is different for Telegram, signal and Whatsapp. Also, in what form do you desire the sentiment analysis report? Shall I create a new file which lists the sentiment score for each message ?

[mohitmishra786]

@JRavi2 As i am working on this can you please put this issue a label of occupied please

[JRavi2]

@asaraf7 this issue has already been assigned to @mohitmishra786, so you need not work on this.