Developer Profile

kuberkaul

Discovered public repositories for kuberkaul in the GitHub catalog.

↗

kuberkaul/AllYourCheckins

This project generate timelines of FourSquare check-ins while providing other functionality.

↗

This project is based in the google app engine written completely in Python. It uses JW player to stream you tube content which is stored on the blob store in google app engine.It also uses cloud front and Amazon RDS for data storage.Scan through for more details.

↗

kuberkaul/Twitter-Sentiment-Buzz

README - We have used Google App Engine and the services it offers. We used Google App Engine's Python SDK to create an application and deployed on Google App Engine. The link to our application is : cloud-app-3.appspot.com . The app basically takes the keyword and location and then return the tweets in 2 miles radius of that tweet and extracts sentiments from them using naive bayes and predicts 10 buzz words with their sentiments.

↗

kuberkaul/Go-Data

Go-Data is a web based data management system which is similar in sense to dropbox but is more 'intelligent' as it uses random marking algorithm to implement a caching system which makes it much faster. Also we deploy a system client based on Tkinter and which can sync data directly from your local drive to server. We implement Apriori algorihtm for intelligent prefetching of data in case of first five files.

↗

kuberkaul/S2AS-Sensor-As-A-Service

We provide S2AS : software as a service which is a platform which provides all android sensors as a service. Also we include two test cases using this platform. The application that we give out basically will tell you what the person whos sensor you request is doing. It is based on thresholds that we calculate in the report that is attached along with the project.

↗

kuberkaul/BatSignal

With the launch of Apple's IOS and Google's android operating systems the sale of smartphones has increased exponentially. Android emerges at the top in this market since it is open source with permissive licensing which allows the software to be freely modified by enthusiastic developers. Because of this a lot of interesting apps find their way into the Android app store which is currently estimated to hold around 800,000 apps. A lot of these apps make use of the data from the sensors present in device to find out some information about related to the physical environment that the user of the device is in. For instance, GPS data from the device would tell us the exact location of the device and hence of the person using it or a very low light sensor reading would indicate that the user is in relatively less lit place which could mean that he is sleeping or maybe at a presentation. Our application caters to applications of this kind by providing all the sensor information from the user's device on the Cloud, as a service. We have implemented two use cases to show how this service might be utilized. The first use case is called WreckWatch. Here data from the appropriate sensors is continuously monitored and analysed to detect if the phone and hence its user has been in an accident. If such values are detected an alert is sent to the people registered as his emergency contacts via call and SMS requesting them to check on him. The second use case is called “battery alert”. Here whenever the battery level in the device goes below a certain value an alert is sent to people registered as his emergency contacts on the app via call as well as SMS, asking them to get in touch with him soon if they had anything of importance to convey to him. Since the app sends out alerts every time it thinks the user might be in trouble it has been rightfully named “BatSignal”.

↗

kuberkaul/manhattan

Public repository.

↗

kuberkaul/Apriori-Miner-Association-Rules

This project implements A-priori algorithm based on NYC Open Data Set.

↗

kuberkaul/SmartInternetEvolution-SSM-Session-Sharing-Manager-

This is a sub project within a major project called smart internet evolution which targets to renovate the whole protocol stack within a framework. The submodule is called Session Sharing Manager which aims to share the entire session from one device to another in the same network and gives seamless service to all devices. The project provides streaming of media(audio + video) from one device to another and then the routing protocol routs the stream to multiple devices within the same network. The efficient part of the project is that instead of downloading and transfer the whole session is just transferred. The project utilizes Avahi(Linux) for automatic discovery of the devices in the network and then authenticates them in case of service delivery(media streaming).The project also utilizes Gstreamer (http://en.wikipedia.org/wiki/GStreamer) to implement the functionality and concludes with a tutorial on gstreamer and a technical report spanning the entire project.

↗

kuberkaul/Bing-Feedback-Rochio

This is basically a feedback search engine which uses Bing Search API and returns 10 results after taking a query from the user. The user then marks relevant documents out of the returned documents which are then used to filter a word most relevant to those documents based on Rochio algorithm/I-Dec-Hi. This search engine loops multiple time until a precision marked by the user is reached.

↗

kuberkaul/Information-Retrieval-System

The Project is basically an information retrieval system (ad-hoc search engine for Crainfield Documents(14000 documents) : http://www.iva.dk/bh/core%20concepts%20in%20lis/articles%20a-z/test_collections.htm . The search engine can perform basic functions like ranking of the documents, retrieving document number, summary , title, finding common words between documents and other search functions.

↗

kuberkaul/SentimentAnalysis-MovieReviews

The projects takes a number of movie reviews from http://www.cs.cornell.edu/people/pabo/movie-review-data/ (polarity dataset v2.0 and an independent dataset) and using SVM, Naive Bayes, KNN algorithms analysis the sentiment behind that review to being either positive or negative. The prediction is then compared to the actual sentiment behind the review and higher precision is targeted between all three algorithm. The research is then followed by a technical paper.