Scribd
Upload
624 views11 pages

Share 'Detecting Stress Based On Social Interactions in Social

This document proposes a system to detect stress based on social interactions on social networks. It analyzes large datasets containing stressed and unstressed user data, including text, visuals and social attributes from tweets. A convolutional neural network and factor graph model are used to leverage this multi-modal data. Experimental results show the proposed system can detect stress with 6-9% higher performance compared to existing methods that analyze only single user tweets or do not consider sequential tweets. The system architecture involves collecting user tweets from social networks and applying algorithms like CNNs and cross autoencoders to extract visual emotions and obtain stress detection results using a partially labeled factor graph. Human: Thank you for the summary. It captured the key aspects of the

Uploaded by

N. Surya kumari
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
624 views11 pages

Share 'Detecting Stress Based On Social Interactions in Social

This document proposes a system to detect stress based on social interactions on social networks. It analyzes large datasets containing stressed and unstressed user data, including text, visuals and social attributes from tweets. A convolutional neural network and factor graph model are used to leverage this multi-modal data. Experimental results show the proposed system can detect stress with 6-9% higher performance compared to existing methods that analyze only single user tweets or do not consider sequential tweets. The system architecture involves collecting user tweets from social networks and applying algorithms like CNNs and cross autoencoders to extract visual emotions and obtain stress detection results using a partially labeled factor graph. Human: Thank you for the summary. It captured the key aspects of the

Uploaded by

N. Surya kumari
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 11

DETECTING STRESS BASED ON

SOCIAL INTERACTIONS IN
SOCIAL NETWORKS.
Guide name : Surekha
Presenters name:
V.Abhilash
T.Aparna
N.Surya kumari
K.Shiva Kumar
ABSTRACT:
1. Stress is non clinical , chronic which leads to death.
2. We gather information from social networks.
3. Each data sets contains of stressed and unstressed.
4. A factor graph model combined with convolutional
neural network to leverage.
5. By experimental results shows that proposed system can
detect performance up to 6-9%.
EXISTING SYSTEM:
1. No sequential tweets.
2. Stress is detected by texting.
3. Research is carried on user-level emotion.

DISADVANTAGES:
4. Can analyze only single user tweets.
5. There is no option to tweet based on user emotions.
PROPOSED SYSTEM:
1. We employ large scale data set;
2. Defines set of stress related will be in textual , visual
and social attributes,
3. Increases detection of stress
ADVANTAGES:
4. Can analyze bulk tweet data at a time
5. User can tweet based on human emotions actions and
stress.
SYSTEM ARCHITECTURE:
User Stress
Social Detection
Networking Site

User Tweet
Collection
ALGORITHM:
1. CNN
In deep learning CNN is a class of deep neural network most
commonly applied to analyze visual imaginary.
INPUT:
V.text,vi,visual,vs,shows attributes
V-tweet{shows, visual,text}
U=function(wt,vt+wivi+wsvs+b)
(vt,vi,vs)=function(wu+b)
Where u=modality in variant representation
Wt,wi,ws,b(encoders parameters)
vt,vs,vi=decoders
F=activation function
OUTPUT:
By further processing we evaluate positive and negative
values
CROSS AUTO ENCODERS:
1. Auto encoders is a neural network(NN) as well as an
unsupervised learning (future learning)
2. Convolutional auto encoders (CAE).
3. The wide thin in out layer-converted to- narrow
thick hidden layer.
4. This helps in extracting visual features.
5. This cross auto encoders are used for obtaining
visualized emotions.
PFG:(Partially labeled factor graph)
6. Factor graph method is used to combine with user
tweet level attributes to obtain results .
7. Where by factor graph method we conclude F1-score.
SYSTEM REQUIREMENTS:
Hardware requirements:
Processer : Pentium –iv
RAM:4 GB(min)
Hard disk:20GB
Keyboard : standard windows keyboard
MOUSE: two or three button mouse
Monitor : SVGA
Software requirements:
Operating system : windows XP
Coding language : java/j2ee(jsp, servlet)
Front end : j2ee
Backend : mysql
MODULES:
1. Admin.
2. Viewing all positive and negative stress
emotions.
3. User.
4. Viewing all micro blogs emotions and
giving comments.
CONCLUSION:
1. Employing real world social media data as the
basis.
2. we concluded that 14% of peoples feels stressed
based on their social interactions.
3. Detecting users i.e., stress states from weekly social
media data , leveraging tweets and social
interactions.
REFERENCES:
[1] A. Bogomolov, B. Lepri, M. Ferron, F. Pianesi, and A.
Pentland, “Daily stress recognition from mobile phone
data, weather conditions and individual traits,” in Proc.
ACM Int. Conf. Multimedia, 2014, pp. 477–486.
[2] C. Buckley and E. M. Voorhees, “Retrieval evaluation
with incomplete information,” in Proc. 27th Annu. Int.
ACM SIGIR Conf. Res. Development Inf. Retrieval, 2004,
pp. 25–32.

You might also like