These abusive comment detection and deletion in live chat application is made during the minor project creation. You can Create your own live recognition and deletion application by taking reference of it.
We worked with many supervised and unsupervised machine learning algorithms such as Naive Bayes, Support Vector machine (SVM) for better result also used rule based approach like afinn, textblob, glove, word2vec which work on Bad or Good classification on data array where 0,1,and negative number makes that decision.
Sentiment Analysis work under Natural Language Processing in which we perform analysis to find out what the sentiment of user has shown in the text form by commenting their view on that particular content published online by someone.
In model creation we also take care of tokenisationn, normalisation, lemmination, steming, CountVectorizer, tfnld, stop word, embeding, nltk courps. We got 56.6% accuracy by using rule based approach and by creating own model, test and training the dataset then we fairly get the accuracy of almost 81.2%.
By referencing many research paper we successfuly managed to create live deletion application which can act as feature for betterment of safe web/mobile platforms for example social media sites such as facebook, instagram and many more where user interaction takes place.
Web site looks like-
- Login Page:
- Sign-Up Page:
- Home Page:
- About-Us Page:
- Contact-Us Page:
- Video Page:
- Before Comment Recognition:
- After Comment Recognition:
Languages we worked on-
-
Python
-
Javascipt
-
Html
-
css
Team members -
-
Rajil Jain
-
pushpraj singh jadhoun