headerphoto

Thesis work

Master's thesis

Proactive personalized mobile information retrieval: I am currently developing feature-rich mobile applications leveraging on user behavioural profiles in collaboration with the Nokia Research Center, Palo Alto. I am setting up a recommender system for RSS feeds employing a hybrid of collaborative and content-based filtering approaches with adaptive user-profile learning from implicit and explicit user feedback. This research is being conducted as an integral part of the Proactive Personalized Information Integration and Retrieval (PIIR) research project at UCSC. Recoo, a system which I designed and developed using this framework, is a recommender system that provides personalized RSS video feeds to mobile phone users from YouTube. This system was deployed as a web application and tested on Nokia N95. We are currently looking at developing a stand-alone application for mobile phones that will also prefetch videos onto the user's mobile phone for offline viewing. We are also exploring possibilities of quicker user-profile learning using active-learning strategies.

Bachelor's dissertation

Design & implementation of reusable user-interface(UI) components: Implemented UI components for web application development, based on JavaServer Faces(JSF) framework in Sun Microsystems, India. [read more]


Selected projects

Graduate

  • Design & Architecture of Recoo: An RSS Recommender system for YouTube feeds

    Abstract: Today's internet users face a bewildering number of choices when looking for information about movies, music, videos, blogs, restaurants, etc. Most websites provide RSS (Really Simple Syndication) feeds, to publish frequently updated content to users on topics in which they are interested. However, the number of RSS feeds and the amount of information channeled through them is increasing exponentially. Recommender systems are considered as a potential and powerful solution to this ubiquitous information overload problem as they offer users a more intelligent and personalized mechanism to seek out new information. In this paper, we describe "Recoo" - a recommender system developed to provide users recommendations of Youtube RSS Feeds. As media content is becoming more and more popular and is being widely used for education as well as entertainment, a system recommending such content will be very useful.
    [Report]

  • Literature survey of Active Learning algorithms

    Abstract: The most time consuming and expensive task in machine learning is the gathering of labeled data to train the model or to estimate its parameters. In the real-world scenario, the availability of labeled data is scarce and we have limited resources to label the abundantly available unlabeled data. Hence it makes sense to pick only the most informative instances from the unlabeled data and request an expert to provide the label for that instance. Active learning algorithms aim at minimizing the amount of labeled data required by strategically selecting the data instance to be labeled by expert. This literature survey aims at providing an insight into the research in this area and categorizes some of the algorithms proposed based on main characteristics.
    [Report]

  • Java-based KNN classifier for text categorization

    Abstract: Categorization of texts into topical categories has gained booming interest over the past few years. There is a growing need for tools that help in finding, filtering and managing the high-dimensional data due to the rapid growth of online information. Building a text classifier by hand is time consuming and costly and hence automated text categorization has gained a lot of importance. A general inductive process automatically builds a classifier by learning, from a set of previously classified documents, the characteristics of one or more categories. In this project we look at the main approaches that have been taken towards text categorization. The K-nearest neighbour algorithm is used for building a classifier for the Reuters-21578 collection.
    [Report]

  • Enriching web browsing experience with Greasemonkey script Firefox extensions

    Abstract: Physical address locations, phone numbers, email addresses etc., can be found virtually on all webpages nowadays. Web email clients, maps, internet phone service applications are some of the most commonly used applications that make use of this information. A user who is looking for directions to a restaurant, searches the web for the address of the restaurant and then copies the address from the restaurant's webpage and uses a map application, say Google maps, to get the directions. If the browser can automatically provides a link to the directions of the restaurant, several intermediate steps of the user are automated. In this project, we have tried to add more functionality to the browser by developing several such useful applications with Greasemonkey scripts that can make the Web user's life easier.
    [Report]

Undergraduate

  • Public access library information management system using Oracle 8 & Visual Basic 6.0
  • Super market management system using C++ & MySQL

Course work

Graduate

Fall 2006

Winter 2007

Spring 2007

Fall 2007

Winter 2008

  • Seminar on Technology and Information Management

Relevant undergraduate courses

  • Artificial Intelligence and Expert Systems
  • Neural Networks and Fuzzy Systems