 Good morning everyone. I am Sushma. I am delighted to be here to present my project, college elections using an upper source software, Lokluck. So, I am Sushma. I already introduced myself so you can just find me at these tags. So, this is my agenda for today. What I am going through briefly. I would be stating about Lokluck and story behind my project development and then about Lokluck API and how I have built the workflow and also the demo of the project and further contributions. So, moving on to next slide. So, as a computer enthusiast, I always dreamed of doing some real-time applications which can contribute to computer society. So, here now today, I am proudly saying that my game has been fulfilled through this project. I had this passion and it always has given me to contribute to open source projects. So, I have been contributing to Lokluck as well as Susie. They are part of Fosatia organization from December 2016. So, Lokluck is a free and open source software which actually provides message search results from Twitter in the JSON format which are very faster and efficient in fetching the data results and also it maintains anonymity. For example, if I want to build something using Twitter data, then I need an OAuth account that is the authentication account. So, on a purpose, if you want to build something, you will not have any anonymity maintaining there. Your credentials, your IP address, everything will be viewed over there. So, there are chances of not getting hidden from the Twitter API. So, for that, Lokluck, what it does is it provides all the interfaces. Like, it actually builds the backend for you. The only thing we need to do is using the APIs, build your friend. And also, it provides us anonymity. So, our IP address will be under disclosure. They won't get to know. So, when I was in my college, very first year, I have seen that every six months we have periodically elections for class representatives. So, for that, we need to elect a leader of our choice and then we have to vote them on a paper. So, I have seen, like, there is a litter. Like, next day when we go after the elections, everywhere in the college there was litter. So, then I found, like, cleanliness precamps. So, then I was like, no, it is not a proper solution. There should be some system so that people can go there, use the portal and then submit the forms. And also, the portal should also help the people to get elected leader very easily. So, for that, what we initially came up with is a website. But the website didn't work that efficiently because we need to maintain the server in our college locally to host that website. And also, we had to maintain the database and it is also a tedious process. So, these kind of disadvantages has made me to think in a proper way. So, I thought there should be something. Definitely, there should be something in order to fetch the results faster at the same time to make the application user-friendly. So, then I approached one of my seniors. She told me that she's contributing to an open source. So, then it was the first time I heard about open source platforms. Then my interest was for Zesha slowly developed and then I started contributing and slowly came up with this project. This is my first project under open source. So, what are the main requirements my college wanted is the application should fetch the data faster depending upon the tweets to declare the results of the election for the class to present it. So, what I wanted is I wanted to use an open source definitely because open source is for contributors like us. It is used by the people like computer enthusiasts like me. So, I had an option of using this open source API. So, this I found look like search API is the best one because it collects huge amount of data at once. It doesn't take too much of time in collecting the data and also we can create our own search portal using this and also we can even derive the statistical evaluation with the data. Once we get the data of aggregations like we have their user names and their count of tweets. So, once we get how many counts of tweets we got then we can use it and evaluate and even we can show it in pictorial manner also. So, when I started playing around with look like APIs initially I was just experimenting each and every API. There were many different categories like search, peer to peer, administration, geolocation, content management, etc. All these APIs are provided by the look like. Depending upon the purpose of your project we can choose one API among these categories. And in that I played around with all the APIs checking with what each API actually does and I found very interesting results. Like all the JSON formats of the API what it collects is from Twitter. And also we are remaining anonymous by collecting the data. So, this made me choose look like API for my college elections app. So, this is the workflow of overall framework like how I have thought of building it. First of all we have college IDs. Like each college has a unique role number for its student and role student. So, as per college norms we have the IDs there. Those IDs have turned them out to be hashtags and use look like search API in order to track those nominated hashtags. And students were announced a time period to tweet. So, during that time period the students they all tweet together depending upon their favorite candidate to be nominated. And then on the projector screen we will show once the time period is done for tweeting. We will show the results like each candidate got how many results like how many counts of tweets. So, which candidate has got more number of tweets they will be declared as the winner. And they will be elected as the class representative. So, this way I have used the look like API to construct my application. So, this is the main workflow. First of all the main GUI screen, the front end screen will be showing the list of candidates which are nominated along with the polls. And while at the back end the look like API what does it work is it query the results from the Twitter directly. And it's like I use search aggregation API here. That is look like dot org slash API slash search dot JSON. Here Q is the query where since this is the time period and source equals to catch for aggregation API. It is must to use source as the catch and then the count and the fields as mentioned here we have the JSON format in that aggregation. I took the mentions because the mentions category list has username along with the tweets, number of tweets. So, the key value pair consists of first key is username and the value is number of tweets. So, that count I used in order to calculate how many tweets overall a hashtag has got that is a candidate has received. So, I have actually hosted this. So, in order to host my application first I have cloned the repo of apps dot look like dot org where all the apps of look like are being deployed. And then to install the server in order to run the look like part npm install hyphen g HTTP server. But before that we need to check whether we have node jays in our system installed. So, then to run the server HTTP server hyphen o it done it do it does the part of displaying the entire applications there. So, these are the related APIs which I have come across and used during my project development. That is the search as an API aggregated the aggregation values giving and then checked my quality of app before hosting it at this particular site. And then the published the app by sending a pull request to app dot look like dot org GitHub repo. So, this has finally done and when I have showed it at my college the impact was profoundly awesome because people didn't expect that local does such miraculous thing because when we hosted a website at my college it was very slow like the search results were very slow. People got exhausted for waiting for the results but look like didn't does the work like it in a matter of seconds it gave me the results. So, people are attracted towards it and even few people like the UX and they found that it is an easy way to elect a candidate. So, this is a representation like demo at my college few of the screenshots. I explained them how does fosatia work actually what are the reports under fosatia what people can contribute and how one can develop his or her own application under look like. And I have even showcased demo of my project in at my campus. So, this is my view of the UI page the front look so I have actually my college has only two nominated candidates. So, I just followed this framework where first candidate and second candidate can be compared and then the number of votes are displayed. And also I used high chart.js as a reference for my pie chart display and finally I declared the winner like whoever got the more votes more number of votes the sum they will be declared as the elected candidate. So, I would like to show demonstrate the live demo now. I actually hosted it on the apps.localoc.org. So, if you go through the website you have many such apps many such students other contributors who have been contributed to it and here's my app. So, for example, now I'm taking fosatia and look like as candidate one and candidate two or anything we can take. So, just let me show with fosatia and look like now. So, even the search results are faster maybe the net is low but when I used my Wi-Fi trust me the results were very fast. So, this is how it is like very interactive in a matter of seconds we are getting the results and at the same time I'm displaying them pictorially which actually gives a interest to users to immediately get to know. Yeah, okay fine she's the winner or he's the winner. So, here from past few days as we all were in fosatia conference we were tweeting about hashtag fosatia. So, fosatia has got many votes and it is the winner. What others can contribute for my application is as it already hosted in the GitHub repository you can go and you can go through the code across and have a look. And if you think that this so and so future is very good to integrate with this application you can contribute and also you can even use it for your own campus or your own local area elections. My ultimate aim for this is making this worldwide like hosting elections online so such kind of thought I got. So, that would be even more like a star in my dream and can style the graphs like in highchart.js reference if we go there are different kinds of charts pie chart bar chart graph chart etc. There are many different kinds of chart. So, we can use any kind of chart once we get the data we can use it for displaying the chart lively which actually impress anyone and attracts their eyes. So, the future goals are that like develop a bar graph for that which constantly shows the improvising results. During the one hour period suppose I give students like I'm giving you one hour time period in that one time one hour time period I need you people to do it to your favorite candidate. So, during that one hour period I'm not switching on this but during that one hour period if I switch on this if I can have a bar chart where the tweets can be sometimes go up sometimes go down because during the one hour periods few people texted first one hour few people at end of the hour. So, that kind of graph which shows increasing and decreasing order that would be more interesting like it can be like a search analysis of the tweets and also to implement it for large scale elections like not just permitting it to minimum levels of a college or campus. So, I have hosted I have told you already the URL for my app is this you can note it down and go across and check please. And also on the same same website it has the GitHub repo you can view my source code you can just go through that and I would be glad if I received any feedback for improvising it also further. So, you can find me at the following links and that is my email ID. So, any further improvisation suggestions I would be very glad and that is my URL link. Thank you so much. It won't go like that there will be the filter check like for one tweet one username like the same username repeats right like as you said if there are multiple tweets on the same with the same user then what does it happen it takes all the user names. So, there will be multiple user names of same category. Suppose I am there I tweeted on Phosasia from past three days like five six tweets but my username is same. So, what does it happen username is unique right. So, how many counts it has it doesn't take if that Phosasia has this particular username then count one that's it. Yeah, you can vote but here we have the we can have the filter that can actually check it won't take the multiple tweets under my account I have some 15 tweets of Phosasia but it takes only one username because 15 user names display the same name. So, that's right. Yeah, yeah. Yeah, the thing is whether it is positive or negative as of now I'm not checking the emotional content here of the tweet so that would be a good addition for add-on for this application actually. So, for that according to now what instructions we pass to students is you have to vote for your favorite candidate. So, even if you vote negatively that would be count on for the other candidate. So, if we say like that they won't do that but still that is an add-on to be implemented here that would be a very good feature. Yeah, actually the Twitter we can't directly delete the tweets over there after this thing right. So, because I used Twitter because local KPI I found very interesting which actually deals with Twitter and also from other sources. So, I have to fetch that what is the other way in order to choose the tweets not directly from Twitter. Yeah, that is what I just mentioned. There will be a filter which actually checks suppose I'm there I tweeted for both local and for Zesha for example then what it does it actually fetches one. So, I already I tweeted for for Zesha once and with my pair like my value my username is already present in local also then it won't accept those both. So, the first one yeah. Only one either of one. Thank you. I will also thank for Zesha for giving me this opportunity. Thank you everyone.