Harsh Jhamtani

Identifying Suggestions for Improvement of Product Features from Online Product Reviews

Online forums are used to share experiences and opinions about products and services. These forums range from review sites such as Amazon (www.amazon.com) to online social networks such as Twitter (www.twitter.com). The user-generated content in these platforms capture the users’ opinions and sentiments. In this work, we explore the problem of identifying suggestions from text content. The paper first defines suggestive intent and then presents a supervised learning approach to identify text that contains suggestive intent. The results show high accuracy with a F1 score of 0.93.

Modeling End-of-Online-Session From Streaming Data

Engagement of consumers has become increasingly important for online marketers. When a potential consumer arrives on its online platform and interacts with it, two important and interrelated questions arise. One whether the consumer is engaged in the session or has completed the session. Two, upon completion of a session whether the consumer will return to the site. Real time answers to both these questions benefit the marketer directly by facilitating more effective retargeting, determination of which is a significant problem in online commerce. We address this problem of retargeting by using automated predictive models. Our model allows a marketer to decide in a real time manner whether a click is the last click of the session. Then the model identifies real time the consumers propensity to return when the session actually ends. This propensity is used to decide whether and whom to retarget with a message. Tests of our model on real data from internet e-commerce sites perform well. The proposed approach is a considerable improvement over the current approach of having to wait for a pre-specified amount of time after a click, in order to identify the end of the session.

Generating Multiple Diverse Summaries

Authors often re-purpose existing content to create shorter versions for other channels. Automatic summarization techniques can be used to generate a candidate content that can be further fine-tuned by the author. Existing work in automatic summarization primarily focus on providing a single succinct summary. However, this may not suit the needs of a content author or curator, who may want to repurpose/select the content from several alternative candidates. In this paper, we propose an approach to generate multiple diverse summaries, so that authors can choose an appropriate summary without compromising on the summary quality. Our approach can be utilized in conjunction with a large class of extractive summarization techniques, and we illustrate our approach with several summarization techniques. We experimentally show that our approach results in fairly diverse summaries, without compromising the quality of the summaries with respect to the single summary generated by the corresponding base methods.

A Supervised Approach for Text Illustration

In this paper we propose a novel method to illustrate text articles with pictures from a tagged repository. Certain types of documents, like news articles, are often accompanied by a few pictures only. Prior works leverage topics or key phrases from the text to suggest relevant pictures. We propose a supervised model based on features like readability, picturability, sentiment polarity, and presence of important phrases, to identify and rank key sentences. The proposed method then suggests some relevant pictures based on the top ranked sentences thus identified.