Arun C. Surendran

Incremental aspect models for mining document streams

In this paper we introduce a novel approach for incrementally building aspect models, and use it to dynamically discover underlying themes from document streams. Using the new approach we present an application which we call “query-line tracking” i.e., we automatically discover and summarize different themes or stories that appear over time, and that relate to a particular query. We present evaluation on news corpora to demonstrate the strength of our method for both query-line tracking, online indexing and clustering.

A Generative-Discriminative Framework using Ensemble Methods for Text-Dependent Speaker Verification

Speaker verification can be treated as a statistical hypothesis testing problem. The most commonly used approach is the likelihood ratio test (LRT), which can be shown to be optimal using the Neymann-Pearson lemma. However, in most practical situations the Neymann-Pearson lemma does not apply. In this paper, we present a more robust approach that makes use of a hybrid generative-discriminative framework for text-dependent speaker verification. Our algorithm makes use of a generative models to learn the characteristics of a speaker and then discriminative models to discriminate between a speaker and an impostor. One of the advantages of the proposed algorithm is that it does not require us to retrain the generative model. The proposed model, on an average, yields 36.41% relative improvement in EER over a LRT.

Data mining and audience intelligence for advertising

Growth in the global advertising industry - especially the recent rapid growth in online advertising - has generated large volumes of data, bringing along with it many challenging data mining problems. Researchers from various disciplines have brought their expertise to solve these exciting problems, leading to a plethora of novel applications and new algorithms. We strongly felt that we needed a forum where data mining researchers and practitioners, from both academia and the industry, could come together to share their experience on advertising. To this end, we organized ADKDD 2007 1, the First International Workshop on Data Mining and Audience Intelligence for Advertising, in conjunction with KDD 2007 at San Jose, California, USA. In this report, we will present a summary of the workshop.

Report on the second KDD workshop on data mining for advertising

Following the success of our first workshop, we organized ADKDD 2008 1 - the second International Workshop on Data Mining and Audience Intelligence for Advertising, in conjunction with KDD 2008 at Las Vegas, Nevada, USA. This report is a summary of the workshop, including brief descriptions of the accepted papers.