Make a bot to decect spam

ScolderCreations

This should exist but perhaps the limits should be set high (once someone made a topic that said this: )

lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele lelelelelelelelelelelelele

And that was a bother to deal with (hard to report)

Last edited by ScolderCreations (Aug. 27, 2022 02:09:02)

charizardsnow

No support st messed up the filter. They would flip this up. This would be very hard to work with.

HeliosTheSuperior

Not sure what it means to “decect” spam or if you meant “dissect” or something, but this sounds dumb. This would cost too much money for something unimportant. I don't think the site should be run by bots, because bots are incorrect often.

k0d3rrr

HeliosTheSuperior wrote:
Not sure what it means to “decect” spam or if you meant “dissect” or something, but this sounds dumb. This would cost too much money for something unimportant. I don't think the site should be run by bots, because bots are incorrect often.

They naturally misspelt “detect”, hence why it says “decect spam”.
But like you said, bots aren't actually intelligent. Even if they “learn” from their past mistakes, the bot will keep on making the same mistakes, and this will result in a lot of false positives and false negatives.

charizardsnow

k0d3rrr wrote:
HeliosTheSuperior wrote:
Not sure what it means to “decect” spam or if you meant “dissect” or something, but this sounds dumb. This would cost too much money for something unimportant. I don't think the site should be run by bots, because bots are incorrect often.
They naturally misspelt “detect”, hence why it says “decect spam”.
But like you said, bots aren't actually intelligent. Even if they “learn” from their past mistakes, the bot will keep on making the same mistakes, and this will result in a lot of false positives and false negatives.

Yes. This would BE HORRIBLE THIS FEATURE. (the bot)

HeliosTheSuperior

k0d3rrr wrote:
HeliosTheSuperior wrote:
Not sure what it means to “decect” spam or if you meant “dissect” or something, but this sounds dumb. This would cost too much money for something unimportant. I don't think the site should be run by bots, because bots are incorrect often.
They naturally misspelt “detect”, hence why it says “decect spam”.

it was joke maybe i shouldve cleared that up lol
just making fun of the glaringly obvious typo in their title

yavuz61035

PGBFLITE6373 wrote:
QuantumScratcher wrote:
typing styles like this will be wrecked by the 500 chars without punctuation filter
500 characters is a lot of characters. It really is a lot. It's smaller than you would think, but it's still really big. Honestly, a forum post that goes over 500 characters without any punctuation is kinda strange. Writing a paragraph without punctuation is not common on the forums. Most paragraphs have to be well thought out and mostly grammatically correct. I don't think adding a random peirod to any of the sentences you say wouldn't be that hard.

wait i just figured out i sometimes use punctuation (such as parantheses and like exclamation marks!)

also this has become an accidental bump

Last edited by yavuz61035 (Sept. 8, 2022 18:28:40)

PGBFLITE6373

bump

blablablahello

does this not already exist?
like if I tried to post “aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa”, the filter bot would stop me because it thinks im spamming

PGBFLITE6373

blablablahello wrote:
does this not already exist?
like if I tried to post “aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa”, the filter bot would stop me because it thinks im spamming

well, you posted it so no?

blablablahello

PGBFLITE6373 wrote:
blablablahello wrote:
snip-
well, you posted it so no?

oh stupid me you meant on the forums
full on support then, anti-spam is probably the only thing the scratch filter can do fine (not good, but not too bad either)

Prince_Wolf1

No support, think of the comment filter, it’s already “stupid” to say the least, a spam filter will be h o r r I b l e
Also a lot of emojis isn’t always bad, and now people can’t say “XDDDDD” probably…

PGBFLITE6373

bump

The_Blue_J

HeliosTheSuperior wrote:
Not sure what it means to “decect” spam or if you meant “dissect” or something, but this sounds dumb. This would cost too much money for something unimportant. I don't think the site should be run by bots, because bots are incorrect often.

It doesn't have to be run by multiple bots, 1 is enough!

PGBFLITE6373

bump.

PaperMarioFan2022

Bump because dupes.

PGBFLITE6373

Bump.

undeterminstic

I have an idea

import re

# Sample function to detect spam
def detect_spam(post_content):
    # Step 1: Check for common spam patterns
    spam_patterns = [
        r"(\bfree\b|\bfreemium\b|\bcheap\b|\bdiscount\b|\bsale\b|\bclick here\b)",
        r"(\bviagra\b|\bcialis\b|\bpharmacy\b)",
        r"(\bwin\b|\bprize\b|\bloan\b)",
        r"(\bmoney\b|\bcash\b|\bearn\b|\bincome\b)",
        r"(\bget rich\b|\benlarge\b|\bweight loss\b)",
    ]

    for pattern in spam_patterns:
        if re.search(pattern, post_content, re.IGNORECASE):
            return True

    # Step 2: Check for excessive links
    # Count the number of URLs in the post content and set a threshold
    url_count = len(re.findall(r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\\(\\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+', post_content))
    if url_count > 3:
        return True

    # Step 3: Check for repetitive characters (e.g., "buyyyy nowww")
    if re.search(r'(\w)\1{2,}', post_content):
        return True

    # Step 4: Check for excessive capitalization
    if len(re.findall(r'[A-Z]', post_content)) > len(post_content) // 3:
        return True

    # If none of the above conditions are met, it's likely not spam
    return False

# Example usage
post_content = "Get a discount on ViAgRa! Click here to win a prize!!!"
if detect_spam(post_content):
    print("This post is likely spam.")
else:
    print("This post is not spam.")

Discuss Scratch