This presentation summarizes a thesis proposal on detecting human emotion on social media based on textual data. The proposal will use a classifier model to identify emotions from social media texts. It will cluster text data into 8 emotion classes to train the classifier. The goal is to analyze social media posts to understand public sentiment on issues and help inform decisions. While the approach only uses text data in English, identifying emotion across languages and media poses challenges.
2. Md. Jeyson Jaman Sawan
Lecturer
Department of Computer Science & Engineering
Supervisor:
Group Member:
Name ID
Md. Jahirul Islam 13334074
Md. Helal Hossain 13334452
4. Using Classifier model to find out human “emotion’’
detection on Social Media based on Textual Content.
AIM OF OUR THESIS
5. What is Emotion?
A natural instinctive state of mind
deriving from one's circumstances,
mood, or relationships with others
6. What is Emotion Detection?
Emotion Detection is the
identification of Human expression
in a dataset.
Identification of a instinctive state
of mind deriving from one's
circumstances, mood, or
relationships with others
7. Cluster the data into groups of different type text
content.
The idea is to treat 8 emotions as 8 different class
for classifier.
Train the classifier with the good training sets and
then go for Testing.
The result of classier will point to a class which is
nothing but a expect emotion.
Classifier -Based
8. Over all accuracy of the model : 71 %
Highest individual class accuracy : 96 %
Why used Classifier –Based?
9. Method is Unsupervised
What type Emotion we will find ?
Validation can be quite challenging(just like for
Classifier)
Finding Emotion in haystack
Emotion Can be by Different language
Challenges of Emotion detection
10. General Steps:
We build an intelligent system .
If you give an input string , our system would
possibly able to say the emotion behind that
textual content.
It will work on English Language Textual
Content.
Emotion detection schemes
11. We want to know about the mental condition of a
people at any national or international issue from
post or comments on social media.
That helps to make a appropriate decision ,what
is good for country or world.
Advantages
12. In the dataset input only textual data.
Image ,pattern , Audio ,video input is invalid.
It will work on only English Language Textual
Content and ignore other language text.
Drawback
13. Finally, we provide information about Emotion
Detection on Social Media based on Textual
Content how we will detect it with Classifier
algorithm that give us a efficiency result.
Conclusion