Nt1310 Unit 3 Assignment 1 Reaction Paper

Fig 3: User Login
Fig 3 shows the user login. It enables the registered user to access the network and allows new user to register the social network.
Fig 4: User Home Page Fig 4 shows the users home page of trustee social network. The short message posted by a user can be shared by other users. Trustee social network provides One Time Password (OTP) recovery by collecting the security details from user. The following Fig 5 shows the OTP security details. The user password is retrieved from bank server of the social network.
Fig 5: OTP Security Details
In this network, the trustees are assigned by individual user as monitors to report on post. The report gathered is to examine whether the sentiment of the post is positive or negative.
When a new user is registered, that account is allotted with trustee level. The post can be shared in public or with friends. The post can be viewed in user wall. The admin processor of social network is preloaded with training dataset for separating subjective word from the post using natural language processing.
The following TABLE 1 describes the attributes that were being used as the training data set and test data in this research work.
TABLE 1: Dataset Description
S.n o
Attribute Name
Attribute Description
1 Month Month Of Published Post
2 Year Year of post published
3 Twitter handle
@ followed by user
4 Display name
Name of the user
5 Tweets No of tweets posted at the end of year
6 TTR Total Time Retweeted
7 EMF End of Month followers
8 TTF Total Time Favorite
The input tweets are preprocessed to extract NLP keywords. The output is collated to real time post to check the sentiment score. When user post diffuse a short message in network, the sentiment level of the post is monitored to calculated the genuine rate. The trustee level of user post decreases with increase in non-genuine rate. The post which attained the minimum trustee level is collated with preprocessed to examine the available genuine rate. If the post is unavailable in keywords extracted using natural language processing then the user who posted the message will be blocked for week of time. The post which is considered as not genuine or rumor will be obstructed to share in network.
Fig 6 shows the notification of obstructed post when other user attempt to diffuse again in network. Fig 6: Notification of Obstructed post
IV.
RESULT
The Sentiment Analysis along with Natural Language Processings obstruct the rumor diffusion in social network. The trustee level created for individual user is monitored. The shared post collate with the dataset keywords to the genuine rate. Depending on the genuine rate the post is permitted to share. In this work, the relevance clustering engaged to rank the relevant keyword. The ranked post with non-genuine factor is removed. This process increase the credibility of the network to diffuse trustworthy information.
V. CONCLUSION
Rumor diffusion is a major problem in social network. Though rumor detection is made with existing datasets. The real time collation results are more efficient than others. There are several algorithms being proposed for detecting rumors, they were not efficient result due to the lack of significance in the field.
But, in this research work, we have built a sentiment analysis and trained it using natural language processing that has resulted in a very high trustee rate. This sentiment rate which is obtained is slightly to be higher than other algorithms that were previously proposed.

[11] Eunsoo Seoa, Prasant Mohapatrab and Tarek
Abdelzaher: