Scribd Logo
Upload

Welcome to Scribd!

  • 296 views

    Instagram Spam Detection Using AI

    Uploaded by

    Dharu Dharani
    The document discusses using AI techniques to build a model to identify spam comments from legitimate comments on Instagram posts by scraping comments, training on a separate spam dataset, and using techniques like N-grams or cosine similarity to classify words and compare comments.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Instagram Spam Detection Using AI

    Uploaded by

    Dharu Dharani
    296 views1 page
    The document discusses using AI techniques to build a model to identify spam comments from legitimate comments on Instagram posts by scraping comments, training on a separate spam dataset, and using techniques like N-grams or cosine similarity to classify words and compare comments.

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses using AI techniques to build a model to identify spam comments from legitimate comments on Instagram posts by scraping comments, training on a separate spam dataset, and using techniques like N-grams or cosine similarity to classify words and compare comments.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    296 views1 page

    Instagram Spam Detection Using AI

    Uploaded by

    Dharu Dharani
    The document discusses using AI techniques to build a model to identify spam comments from legitimate comments on Instagram posts by scraping comments, training on a separate spam dataset, and using techniques like N-grams or cosine similarity to classify words and compare comments.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    You are on page 1of 1

    Instagram Spam Detection Using AI

    Have you ever received a notification that someone commented on your Instagram
    post? You excitedly pick up your phone and open the app only to find that it’s a bot
    promoting some knockoff brand of shoes. The comment section of many Instagram
    posts is filled with bots. They can range from annoying to dangerous, depending on
    the type of call to action they require from you.

    You can build a spam detection model using AI techniques to identify the difference
    between spam and legitimate comments.

    You might not be able to find a dataset that has a collection of Instagram spam
    comments, but you can collect the data for this analysis by scraping the web. Access
    the Instagram API with Python to get unlabelled comments from Instagram.

    You can use a different set of data for training, like Kaggle’s YouTube spam collection
    dataset. Then, use keywords to classify words that commonly appear in spam
    comments.

    Use a technique like N-Gram to assign weightage to words that tend to appear in
    spam comments, then compare those words with each scraped comment from the
    web. Another approach you can take is the use of a distance-based algorithm
    like cosine similarity. These approaches will yield better results based on the type of
    pre-processing you apply.

    If you remove stop-words, whitespaces, punctuation and clean the data correctly,
    you will find that the algorithm performs better as it can match similar words with
    each other.

    You can also use a pre-trained model like ALBERT for better results. While distance or
    weightage matching algorithms work well in finding similar words, they are unable to
    grasp the context of a sentence.

    CORE TERM: Artificial Intelligence – Insta Spam detection

    Team Members:
    R. Dharani
    P. A. Bharathi
    B. M. Nithyashri

    You might also like