Final Year PPT on Twitter App

TWITANCE

Guided By:Prof.Geeta Patil

Sponsored By: Group Members:Josh Software Ashish Singh

Rohit Khatana Jitendra Sharma Aravind S

TWITANCE

● Recommendation Tool● Twitter analytics● Checks the response on tweets● Uses this response to better engage with the

audience

TWITANCEBased On Twitter

Why This?

Whenever we tweet on twitter its always good to have an idea of the impact that tweet had and also the activity around that tweet so as to better engage with interested audience.

Because, people like to know their influence in social media, which will help them to plan their strategies in case of brands and help people to focus more on them.

Use -cases

• Martha tweets about the cooking very frequently.

• If she has a new recipe for a dish and she tweets about it.

• She wants to know response(RT,reply,favorite) on that tweet.

• So as to better engage with people who are more interested in her recipes.

• Doing so its more likely that interested people may come up with suggestions and Rts eventually leading to wider reach.

Impact of promoter● Very recently chennai express was released

and there was a lot of activities around the #chennaiexp and @iamsrk.

● Now if we plan to tweet about the movie. Mentioning @iamsrk increases the possibility of that tweet being retweeted by @iamsrk. Thus increasing the reach of your tweet.

• Normally people don't know who his promoter would be, while tweeting related to a particular topic.

• Twitance will help them get to a decision regarding this.

ANOTHER USE-CASE

● Let's suppose someone wants to survey about the rape-cases in India . And he tweets about it.

● Then on the basis of tweet's impact in India .We can give him the best person who can help him out when he comes to India.

Similar Apps which we found During over research

TWEET REACH:• Processes your tweets and looks for audience and

possible reach , mentions the list of contributors

KLOUT:• Measures the size of your social media network.

TWEET ANALYZER:• Provides influential score on the basis of popularity

TWIST:• Offers Trends on Twitter on the basis of

keywords and hash tag.

TWITTURLY:• Keeps track of popular URLs on twitter.• Works similar to Digg.

Milestones

With Twitance our main aim is to finally be able to show a list of twitter users to the authorized user as recommendations and analytics of a particular tweet. Thus based on the analysis, twitter user can get a fair idea of their reach through tweets.

Our Main idea has these many components:-

• Two lists of recommendations for a particular tweet: Universal List

Followers List

• First a Graph(a pictorial representation) that will show the reach and chains as shown in the picture below.

• So the basic idea is all the shares from Gautam’s tweet will lie in his circle and add to his contribution in the publicity of this post.

TWEET ANALYTICS

2. Reach over timeIn this reach over time is shown this means which users contributed to the post over time and their influence.Important contributors are highlighted.

3. Influencers and other statsThis shows the list of people that influenced the post most this required some math.

Gender and Location wise stats will also be there.

Technologies Used:

• Ruby On Rails:

Open source web application frame Model view controller design principles Active Record Pattern Encourages developers to use restful routes Availability of numerous gems(libraries) makes your

life easier.

• Mongo DB:

Cross platform document oriented database system

No SQL (scheme-less) Supports restful operations (CRUD) Favors JSON like documents with dynamic

schemas.

• Elastic Search [Analyzers + Efficient Searching ]

Wrapper on Lucene project. Open source , distributed server side search

engine. Consists various analyzers like Snowball

anaylzer. Provides an approximate result if actual result

not found.

• Twitter API :

Provides JSON type API. Provides services strictly to the authorized

users. Provides all public data related to twitter

users which includes Followers, Friends, Tweets and public info of users.

Search Rate limit : 180/user/15 min window User Rate limit : 15/user/15 min window

REFERENCES:• IEEE PAPER :Credibility in Context: An Analysis of Feature Distributions in Twitter

John O’Donovan, Byungkyu Kang, Greg Meyer, Tobias Hollerer University of California,Santa Barbara, USA

• www.ruby.org• www.elasticsearch.org• www.rubyonrails.in

THANK YOU!!

Made By :-Rohit KhatanaAravind SJitendra SharmaAshish Singh