Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Stress Detection of User in Social Networks Based on
Social Interactions 1Chinta Someswara Rao,
2K V S S R Murthy,
3V MNSSVKR Gupta and
4R. Shiva Shankar
1Department of CSE,
S.R.K.R Engineering College,
Bhimavaram, W.G. District, A.P. India.
[email protected] 2Department of CSE,
S.R.K.R Engineering College,
Bhimavaram, W.G. District, A.P. India. 3Department of CSE,
S.R.K.R Engineering College,
Bhimavaram, W.G. District, A.P. India. 4Department of CSE,
S.R.K.R Engineering College,
Bhimavaram, W.G. District, A.P. India.
Abstract Mental pressure is debilitating individuals' wellbeing. It is non-
insignificant to identify pressure opportune for proactive care. With the
notoriety of web based life, individuals are accustomed to imparting their
every day exercises and associating to companions via web-based
networking media stages, making it practical to use online interpersonal
organization information for push discovery. In this paper, we find that
clients push state is firmly identified with that of his/her companions in
web based life, and we utilize a substantial scale dataset from true social
stages to efficiently think about the relationship of client's pressure states
and social associations. we dissecting the social association information,
additionally find a few interesting wonders, i.e. the quantity of social
structures of meager associations of focused on clients.
Key Words:Social network, Psychological stress, social interactions, data
International Journal of Pure and Applied MathematicsVolume 119 No. 18 2018, 225-236ISSN: 1314-3395 (on-line version)url: http://www.acadpubl.eu/hub/Special Issue http://www.acadpubl.eu/hub/
225
mining; machine learning.
1. Introduction
For the most part, information mining (now and then called information or
learning revelation) is the way toward breaking down information from
alternate points of view and abridging it into helpful data - data that can be
utilized to build income, cuts costs, or both. Information mining programming
is one of various scientific instruments for breaking down information[1,2]. It
enables clients to break down information from a wide range of measurements
or points, arrange it, and abridge the connections distinguished. Actually,
information mining is the way toward discovering connections or examples
among many fields in substantial social databases[3,4].
2. Literature Survey
Ben Verhoeven et al., [1] chipped away at "A multilingual twitter stylometry
corpus for sex and identity profiling", which identifies identity characteristics of
writers in light of composing style. A few identity typologies exist, the Myers-
Briggs Type Indicator (MBTI) is especially famous in the non-academic
network, and numerous individuals utilize it to examine their own identity and
discuss the outcomes on the web. In this manner, a lot of self surveyed
information on MBTI are promptly accessible via web-based networking media
stages, for example, Twitter.
Golnoosh Farnadi et al., [2] investigated towards "Computational identity
acknowledgment in web based life" about an assortment of ways to deal with
naturally surmise clients' identity from their client produced content in online
life. Methodologies vary as far as the machine learning calculations and the
capabilities utilized, kind of used impression, and the internet based life
condition used to gather the information.
Quan Guo et al., [3] has created "Learning hearty uniform highlights for cross-
media social information by utilizing cross auto encoders". Cross-media
examination misuses social information with various modalities from numerous
sources all the while and synergistically to find learning and better comprehend
the world. While customary element learning techniques center around
managing single methodology information or information combined over
numerous modalities.
Novak P Kralj et al., [4] created "Estimation of emoticons" and talked about
another age of emojis, called emoticons that is progressively being utilized as a
part of versatile correspondences and online networking. In the previous two
years, more than ten billion emoticons were utilized on Twitter. Emoticons are
Unicode realistic images, utilized as shorthand to express ideas and thoughts.
The feeling examination of the emoticons enables us to make a few fascinating
inferences. The feeling dispersion of the tweets with and without emoticons is
International Journal of Pure and Applied Mathematics Special Issue
226
altogether unique.
H. Lin et al., [5] made their work on "Mental pressure location from cross-
media microblog information utilizing profound inadequate neural system"
which demonstrates customary mental pressure recognition generally depends
on the dynamic individual support, which influences the discovery to work
expending, time-costing and hysteretic. With the quick improvement of
informal communities, individuals turn out to be increasingly ready to share
states of mind by means of microblog stages.
H. Lin et al., [6] has contributed on "Client level mental pressure identification
from internet based life utilizing profound neural system", which suggests that
current pressure discovery techniques for the most part depend on mental scales
or physiological gadgets, making the location entangled and expensive. In this
task, first investigate to naturally distinguish person's mental pressure by means
of web-based social networking.
Liqiang Nie et al., [7] chipped away at "Connecting the vocabulary hole
between wellbeing searchers and medicinal services information" giving
insights about the vocabulary hole between wellbeing searchers, the cross-
framework operability and the between client reusability. This paper exhibits a
novel plan to code the medicinal records by mutually using neighborhood
mining and worldwide learning approaches.
Jian Bo Yang et al., [8] contributed "Profound convolutional neural systems on
multichannel time arrangement for human movement acknowledgment". They
centers around human action acknowledgment (HAR) issue, in which inputs are
multichannel time arrangement signals gained from an arrangement of body
worn inertial sensors and yields are predefined human exercises. This technique
embraces a profound convolutional neural system (CNN) to robotize include
gaining from the crude contributions to a deliberate way. Through the profound
engineering, the educated highlights are esteemed as the more elevated amount
theoretical portrayal of low level crude time arrangement signals.
Qian Zhang and Bruno Goncalves [9] examined on "Topical contrasts between
chinese dialect twitter and sina weibo". In this investigation, They differentiate
the discourses happening on Sina Weibo and on Chinese dialect Twitter so as to
watch two unique strands of Chinese culture. A basic specially appointed
calculation is proposed to distinguish themes of Tweets and Weibos.
Xiaojun Chang et al., [10] chipped away at "Semantic idea disclosure for vast
scale zero-shot occasion discovery". They concentrated on distinguishing
complex occasions in unconstrained Internet recordings. While most existing
works depend on the plenitude of marked preparing information, They
considered a more troublesome zero-shot setting where no preparation
information is provided.
International Journal of Pure and Applied Mathematics Special Issue
227
3. System Architecture
The proposed framework includes administrator, administrator server, OSN
client and web database. In this framework, the administrator deals with every
one of the information. The OSN client registers to be a piece of social
associations. The administrator sees client points of interest, acknowledges the
information and approves the OSN client.
The OSN client looks for companions and demands them. They can see every
one of their companions.
The clients can make tweets by tweet name, tweet depiction, tweet picture and
tweet date. They can see all their made tweets and discover positive, negative
and focused on feelings on their tweets. They can see all their companion's
tweets and can retweet.
The administrator sees all companion solicitations and reactions. Administrator
can include tweet classification like positive, negative and focused on feelings.
Administrator can choose tweet classification and include tweet channel and
rundown all channels beneath.
He can list all tweets smaller scale blog with its client points of interest. He can
see positive, negative and focused on feeling tweets. He can likewise see add up
to tweets and rundown of pursuit history. The administrator additionally keeps
up the aggregate site.
We likewise utilize database for putting away and recovery motivation behind
the client subtle elements and other tweet points of interest through which
graphical investigation of the enthusiastic tweets by the clients can be found.
The system structure is shown in Fig 1.
Algorithm Explanation
Input: A series of time-varying attribute augmented network with stress states on
some of the user
nodes.
Output: Graphical Analysis of positive, negative and stressed emotion tweets.
Step 1: Admin adds emotion categories and adds filters to those categories.
Step 2: New users have to register themselves and could only login into their
accounts if admin authorizes them.
Step 3: Users send friend requests and accept other user’s requests.
Step 4: Users post tweets and their friend’s can re-tweet them.
Step 5: Users can also view their friend’s tweets.
Step 6: Admin categorizes the tweets and re-tweets into positive, negative and
stressed tweets.
Step 7: Admin provides the emotional analysis using a graph.
International Journal of Pure and Applied Mathematics Special Issue
228
Fig 1: System Architecture
Modules
Admin: In this module, the Admin has to login by using valid user name and
password. After login successful he can do some operations such as View all
End Users and Authorize, View all friend request and Response, Add Tweet
Category like Positive, Negative, Stressed, Select Tweet Category and Add
Tweet Filter and list all filters below, List all Tweets micro-blog with its user
details, View Positive (+)Emotion Tweets Emotions ,View negative (-)Emotion
Tweet Emotions ,View Stress Emotion Tweets, View total tweets and find
number of positive, negative and stressed tweets ,List of search history, Find
Admin Admin Server
OSN User
WEB
Database
Accepting user information
View user data information
Associate the admin
Process all user queries
User registration
Register and Login
View your profile
Search friends and request friends
View all your friends
Create tweet by tweet name, description, image and
date
View all your created tweets and find positive,
negative and stressed emotions
View all friend tweets and retweets by feeding your
sentiments or comments
Possibilities
View all the end users and
authorize
View all friend requests and
response
Add Tweet category like
positive, negative, stressed
Select tweet category and add
tweet filter and list all filters
below
List all tweets micro-blog with
its user details
View +, - and stress emotion
reviews and tweets
View total tweets and find
no.of +, - and stressed tweets
List of search history
po
ssibilities
International Journal of Pure and Applied Mathematics Special Issue
229
No. Of positive or negative or stressed Tweets emotion in chart
Viewing All Positive and Negative, Stress Emotions: In this module, the
admin can see all Positive and Negative Emotions posted by all users for cross
domain Micro Blogs. The Review is considered either as Positive or Negative
based on the Positive and Negative list of words which are used to find the
review as positive or negative. In this the Positive and Negative words will be
highlighted in blue color and in italics style.
View and Authorize Users: In this module, the admin can view the list of users
who all registered. In this, the admin can view the user’s details such as, user
name, email, address and admin authorizes the users.
View Chart Results: In this, the cross domain number of positive and negative
Emotions for particular post will be shown in a chart by selecting particular
Micro Blogs from a combo box.
User: In this module, there are n numbers of users are present. User should
register before doing any operations. Once user registers, their details will be
stored to the database. After registration successful, he has to login by using
authorized user name and password. Once Login is successful user will do some
operations like View your profile, Search Friends and Request, Friend. View all
Your Friends, Create Tweet by Tweet name, Tweet description, Tweet Image,
Tweet date, View all your created Tweets and find posing, Stress emotions on
your Tweets, View all your friends tweets and retweet by feeding your
sentiments or comment
Viewing All Micro Blogs Emotions and give Comment: In this, the user can
view all the Micro Blogs details, Emotions and user can comment on them by
entering their own Emotions. Each time the rank will be incremented for
particular Micro Blogs once the review is posted.
we use different tables used in our implementation. Table 1 represents for
categories, Table 2 represents for filter, Table 3 represents for friend_search,
Table 4 represents for negative, Table 5 represents for positive, Table 6
represents for req_res, Table 7 represents for retweets, Table 8 represents for
server, Table 9 represents for stressed, Table 10 represents for tweets, Table 11
represents for user structures respectively.
Table 1 categories
Field Type Constraint
si_no int(11) not null
categorie varchar2(50) Null
International Journal of Pure and Applied Mathematics Special Issue
230
Table 2 filter
Field Type Constraint
categorie Text null
filter Text null
Table 3 friend_search
Field Type Constraint
id int(11) not null
username Text null
keyword Text null
date Text Null
Table 4 negative
Field Type Constraint
id int(11) not null
tname Text null
totaluser Text null
negative Text null
Table 5 positive
Field Type Constraint
id int(11) not null
tname Text null
totaluser Text null
postive Text Null
Table 6 req_res
Field Type Constraint
id int(11) not null
requestfrom Text null
requestto Text null
status Text null
dt Text null
Table 7 retweets
Field Type Constraint
id int(11) not null
tname Text null
t_user Text null
t_comment Text Null
r_user Text Null
r_tweet Text Null
date Text Null
International Journal of Pure and Applied Mathematics Special Issue
231
Table 8 server
Field Type Constraint
id int(11) not null
username Text not null
password Text not null
Table 9 stressed
Field Type Constraint
id int(11) not null
tname Text Null
total user Text Null
stressed Text Null
Table 10 tweets
Field Type Constraint
id int(11) not null
tweeter Text Null
tname Text Null
image Blob Null
description Text Null
date Text Null
rank int(11) Null
Table 11 user
Field Type Constraint
id int(11) not null
username Text not null
password Text not null
email Text not null
mobile Text not null
dob Text Null
gender Text Null
address Text Null
pincode Text Null
network Text Null
image Blob Null
status Text Null
Sample registered user details are shown in Table 12.
International Journal of Pure and Applied Mathematics Special Issue
232
Table 12 Registered User Details
Usernam
e
Email Mobile Usernam
e
Email Mobile
1. manas [email protected] 978909089
0
Chiku [email protected] XXXXXXXXX
X
2. mounika [email protected]
m
934567823
1
madhavi [email protected] XXXXXXXXX
X
3. sahithi [email protected] 879067516
4
padma [email protected] XXXXXXXXX
X
4. ashok [email protected] 908453287
6
rajesh [email protected] XXXXXXXXX
X
5. satwika [email protected] 908990876
0
madhava
n
m
XXXXXXXXX
X
6. satish [email protected] 987687690
9
raviteja [email protected] XXXXXXXXX
X
7. tejaswini [email protected]
m
879067895
6
prakashra
j
m
XXXXXXXXX
X
8. Akhil [email protected] 789678567
0
Ali [email protected] XXXXXXXXX
X
9. haneesha [email protected]
m
909897989
7
Sunil [email protected] XXXXXXXXX
X
10. Sai [email protected] 897067896
4
ravibabu [email protected] XXXXXXXXX
X
4. Results
In this section, we discuss the results of the system. We define the positive,
negative and stressed emotions based on the social interactions in social
networks by the users. We make a count of the total positive, negative and
stressed emotion tweets and retweets made against that tweet. The positive
tweets data graphically represented which is shown in Fig 2. From the graph it
is observed that, If the count of the positive tweets is high, then the user is in a
positive state. This could be viewed by his positive emotional analysis.
Fig 2: Postive tweet rank graph
International Journal of Pure and Applied Mathematics Special Issue
233
The negative tweets data graphically represented which is shown in Fig 3. From
the graph it is observed that, If the count of the negative tweets is high, then the
user is in a negative state. This could be viewed by his negative emotional
analysis.
Fig 3: Negative tweet rank graph
The positive tweets data graphically represented which is shown in Fig 4. From
the graph it is observed that, If the count of the negative tweets is high, then the
user is in a negative state. This could be viewed by his negative emotional
analysis.
Fig 4: Stressed tweet rank graph
5. Conclusions
In this paper, we presented a framework for detecting social network people
stress from their yearly, monthly, weekly and daily data. In this paper we
consider negative, positive and stressed tweets and given these data as input to
our proposed approach. Our framework will calculates the correlation among
the posted data, separates the tweets by calculating their rank and this study will
helpful for future estimation of employee or person's behavior.
International Journal of Pure and Applied Mathematics Special Issue
234
References
[1] Ben Verhoeven, Walter Daelemans, and Barbara Plank. Twisty: A multilingual twitter stylometry corpus for gender and personality profiling. In Proceedings of the Tenth International Conference on Language Resources and Evaluation LREC, pages 1632–1637, 2016.
[2] Golnoosh Farnadi, Geetha Sitaraman, Shanu Sushmita, Fabio Celli, Michal Kosinski, David Stillwell, Sergio Davalos, Marie Francine Moens, and Martine De Cock. Computational personality recognition in social media. User Modeling and User- Adapted Interaction, pages 1–34, 2016.
[3] Quan Guo, Jia Jia, Guangyao Shen, Lei Zhang, Lianhong Cai, and Zhang Yi. Learning robust uniform features for cross-media social data by using cross auto encoders. Knowledge Based System, 102:64– 75, 2016.
[4] Novak P Kralj, J Smailovi, B Sluban, and I Mozeti. Sentiment of emojis. Plos One, 10(12), 2015
[5] H. Lin, J. Jia, Q. Guo, Y. Xue, J. Huang, L. Cai, and L. Feng. Psychological stress detection from cross-media microblog data using deep sparse neural network. In proceedings of IEEE International Conference on Multimedia & Expo, 2014.
[6] H. Lin, J. Jia, Q. Guo, Y. Xue, Q. Li, J Huang, L. Cai, and L. Feng. User-level psychological stress detection from social media using deep neural network. In Proceedings of ACM Int. Conference on Multimedia, 2014.
[7] Liqiang Nie, Yi-Liang Zhao, Mohammad Akbari, Jialie Shen, and Tat-Seng Chua. Bridging the vocabulary gap between health seekers and healthcare knowledge. Knowledge and Data Engineering, IEEE Transactions on, 27(2):396–409, 2015.
[8] Jian Bo Yang, Minh Nhut Nguyen, Phyo Phyo San, Xiao Li Li, and Shonali Krishnaswamy. Deep convolutional neural networks on multichannel time series for human activity recognition. In Proceedings of International Joint Conference on Artificial Intelligence, pages 3995–4001, 2015.
[9] Qian Zhang and Bruno Goncalves. Topical differences between chinese language twitter and sina weibo. Computer Science, 2015.
[10] Xiaojun Chang, Yi Yang, Alexander G Hauptmann, Eric P Xing, and Yao-Liang Yu. Semantic concept discovery for large-scale zero-shot event detection. In Proceedings of International Joint Conference on Artificial Intelligence, pages 2234–2240, 2015.
International Journal of Pure and Applied Mathematics Special Issue
235