Abhimanyu Lad

Update:

In April 2011, I joined LinkedIn as a Data Scientist.

Welcome!

I am a fifth year PhD student at the Language Technologies Institute, advised by Prof Yiming Yang. Currently, I am working on a probabilistic framework for modeling retrieval performance in terms of non-independent utility of documents, measured with respect to relevance and novelty of information across one or more ranked lists. This framework would enable automatic evaluation as well as optimization of novelty and diversity-based retrieval systems under more realistic conditions.

Thanks to Yahoo! for supporting my research through the Yahoo! PhD Fellowship for 2007-09.

PhD Dissertation: A Framework for Evaluation and Optimization of Relevance and Novelty-based Retrieval

Research Interests

Information retrieval and statistical machine learning: Novelty and diversity-based retrieval over web documents and news streams, adaptive filtering, probabilistic topic modeling, active learning, multi-task learning.

Top words that appear in my publications:

Publications

• A. Lad, Y. Yang. Learning to Rank Relevant and Novel Documents through User Feedback. In Proceedings of the 19th ACM International Conference on Information and Knowledge Management (CIKM 2010).
• A. Lad, Y. Yang. Active Ordering of Interactive Prediction Tasks. In Proceedings of the 10th SIAM International Conference on Data Mining (SDM 2010).
• Y. Yang, A. Lad. Modeling Expected Utility of Multi-session Information Distillation. In Proceedings of the 2nd International Conference on the Theory of Information Retrieval (ICTIR 2009).
• A. Lad, Y. Yang, R. Ghani, B. Kisiel. Toward Optimal Ordering of Prediction Tasks. In Proceedings of the 9th SIAM International Conference on Data Mining (SDM 2009).
• K. Salomatin, Y. Yang, A. Lad. Multi-field Correlated Topic Modeling. In Proceedings of the 9th SIAM International Conference on Data Mining (SDM 2009).
• Y. Yang, A. Lad, H. Shu, B. Kisiel, C. Cumby, R. Ghani, K. Probst. Graph Structure Learning for Task Ordering. In Proceedings of the 11th International Conference on Enterprise Information Systems (ICEIS 2009).
• A. Lad, Y. Yang. Generalizing from Relevance Feedback using Named Entity Wildcards. In Proceedings of the 16th ACM International Conference on Information and Knowledge Management (CIKM 2007).
• Y. Yang, A. Lad, N. Lao, A. Harpale, B. Kisiel, M. Rogati. Utility-Based Information Distillation over Temporally Sequenced Documents. In Proceedings of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2007).
• A. Lad, M. Saggar. Carnivore - A Web Clustering Program. In Proceedings of IEEE International Conference on Information Reuse and Integration (IRI 2004).
• A. Lad. SpamNet - Spam Detection using PCA and Neural Networks. In Lecture Notes in Computer Science 2004, pp. 205-213 (LNCS 2004).

Book Chapters

Y. Yang, A. Lad. Utility-Based Information Distillation. Book Chapter in Text Mining: Classification, Clustering, and Applications (Chapman & Hall 2009).

Software

• Mekano: A Python library for rapid prototyping and experimentation in information retrieval and machine learning.
• An applet showing decision surfaces created by kNN.
• An applet showing EM for estimating Gaussian mixtures.

Other Interests

• Python -- An excellent programming language for researchers!
  • IPython (an interactive shell for Python)
  • Cython (a tool/language for writing C/C++ extensions for Python)
  • NumPy (a package for scientific computing in Python)
  • matplotlib (a package for 2D plotting in Python, works great with IPython)
• Playing the Tabla
• Digital photography and HDR imaging