At Opinosis Analytics, our expertise spans AI strategy, natural language processing, LLMs, and AI agent development. The publications below showcase years of applied research and innovation that form the foundation of our consulting work today.
The Business Case for AI: A Leader’s Guide to AI Strategies, Best Practices & Real-World Applications
Kavita Ganesan (Book)
Citation: Ganesan, K. The Business Case for AI: A Leader’s Guide to AI Strategies, Best Practices & Real-World Applications. Opinosis Analytics Publishing, 2022.
This book provides practical guidance for leaders on how to implement artificial intelligence responsibly and effectively. It outlines strategies for identifying opportunities, setting success metrics, and managing risks, supported by case studies across industries.
ROUGE 2.0: Updated and Improved Measures for Evaluation of Summarization Tasks
Kavita Ganesan (Research Paper)
Citation: Ganesan, K. “ROUGE 2.0: Updated and Improved Measures for Evaluation of Summarization Tasks.” arXiv:1803.01937, 2018.
Improvement over the widely used ROUGE metric for automatic summarization evaluation. The updated measures capture synonyms and topics more effectively, leading to more reliable comparisons between summarization systems.
Discovering Related Clinical Concepts Using Large Amounts of Clinical Notes
Kavita Ganesan, S. Lloyd, V. Sarkar
Citation: Ganesan, K., Lloyd, S., Sarkar, V. “Discovering Related Clinical Concepts Using Large Amounts of Clinical Notes.” Biomedical Engineering and Computational Biology, 2016.
Methods to uncover semantically related concepts from large collections of clinical notes. The study highlights how text mining can enrich clinical search and medical knowledge discovery.
Linguistic Understanding of Complaints and Praises in User Reviews
Kavita Ganesan, G. Zhou
Citation: Ganesan, K., Zhou, G. “Linguistic Understanding of Complaints and Praises in User Reviews.” NAACL-HLT, 2016.
This paper analyzes the linguistic structures that distinguish complaints from praises in online reviews. It provides a framework for more precise sentiment categorization.
OpinoFetch: A Practical and Efficient Approach to Collecting Opinions on Arbitrary Entities
Kavita Ganesan, C. Zhai
Citation: Ganesan, K., Zhai, C. “OpinoFetch: A Practical and Efficient Approach to Collecting Opinions on Arbitrary Entities.” Information Retrieval Journal 18(6), 2015.
This work introduces OpinoFetch, a scalable system for harvesting opinions on diverse entities from the web. It demonstrates efficient methods for large-scale opinion retrieval.
A General Supervised Approach to Segmentation of Clinical Texts
Kavita Ganesan, M. Subotin
Citation: Ganesan, K., Subotin, M. “A General Supervised Approach to Segmentation of Clinical Texts.” IEEE Big Data, 2014.
A supervised method for segmenting long clinical documents into coherent sections, improving the usability of electronic health records.
Opinion-Driven Decision Support System
Kavita Ganesan
Citation: Ganesan, K. Opinion-Driven Decision Support System. Ph.D. Dissertation, University of Illinois Urbana-Champaign, 2013.
Dr. Ganesan’s dissertation introduces a complete framework for capturing, summarizing, and organizing opinions into structured decision-support tools.
Findilike: A Preference-Driven Entity Search Engine
Kavita Ganesan, C. Zhai
Citation: Ganesan, K., Zhai, C. “Findilike: A Preference-Driven Entity Search Engine for Evaluating Entity Retrieval and Opinion Summarization.” Living Labs for IR Evaluation Workshop (WWW), 2013.
Findilike is a prototype search engine that integrates user preferences with entity retrieval and opinion summarization to deliver personalized results.
Micropinion Generation: An Unsupervised Approach to Generating Ultra-Concise Summaries of Opinions
Kavita Ganesan, C. Zhai, E. Viegas
Citation: Ganesan, K., Zhai, C., Viegas, E. “Micropinion Generation: An Unsupervised Approach to Generating Ultra-Concise Summaries of Opinions.” WWW, 2012.
This work proposes a method to automatically generate tweet-sized opinion summaries, making large volumes of feedback easier to digest.
Opinion-Based Entity Ranking
Kavita Ganesan, C. Zhai
Citation: Ganesan, K., Zhai, C. “Opinion-Based Entity Ranking.” Information Retrieval 15(2), 2012.
This body of work introduces ranking methods based on aggregated opinion signals, providing a sentiment-driven way of ordering entities such as products or services.
Comprehensive Review of Opinion Summarization
H. D. Kim, Kavita Ganesan, et al.
Citation: Kim, H. D., Ganesan, K., et al. “Comprehensive Review of Opinion Summarization.” University of Illinois, 2011.
A survey of techniques for opinion summarization, covering both extractive and abstractive methods and their comparative strengths.
A Graph-Based Approach to Abstractive Summarization of Highly Redundant Opinions
Kavita Ganesan, C. Zhai, J. Han
Citation: Ganesan, K., Zhai, C., Han, J. “A Graph-Based Approach to Abstractive Summarization of Highly Redundant Opinions.” COLING, 2010.
A graph algorithm for condensing large amounts of repetitive opinion text into concise, abstractive summaries.
Finding Related Entities by Retrieving Relations
V. Vydiswaran, Kavita Ganesan, et al.
Citation: Vydiswaran, V. G., Ganesan, K., et al. “Finding Related Entities by Retrieving Relations: UIUC at TREC 2009 Entity Track.” TREC, 2009.
This paper investigates techniques for retrieving related entities using relation-based evidence, evaluated in the TREC Entity Track.
Term Proximity and Normalization in Pseudo-Relevance Feedback
Y. Lv, Kavita Ganesan, et al.
Citation: Lv, Y., Ganesan, K., et al. “Term Proximity & Normalization in Pseudo-Relevance Feedback: UIUC at TREC 2009 Million Query Track.” TREC, 2009.
Exploring the effect of term proximity and normalization in feedback models, aiming to improve search effectiveness in large-scale tasks.
Mining Tag Clouds and Emoticons Behind Community Feedback
Kavita A. Ganesan, N. Sundaresan, H. Deo
Citation: Ganesan, K. A., Sundaresan, N., Deo, H. “Mining Tag Clouds and Emoticons Behind Community Feedback.” WWW, 2008.
Analyzes tags and emoticons in community feedback to extract sentiment and identify themes in user-generated content.
Multi-Factor Clustering for a Marketplace Search Interface
N. Sundaresan, Kavita A. Ganesan, R. Grandhi
Citation: Sundaresan, N., Ganesan, K. A., Grandhi, R. “Multi-Factor Clustering for a Marketplace Search Interface.” WWW, 2007.
Presents a clustering method that organizes search results by multiple factors, improving marketplace navigation.
EntityBases: Compiling, Organizing and Querying Massive Entity Repositories
C. Knoblock, J. Ambite, Kavita A. Ganesan, et al.
Citation: Knoblock, C., Ambite, J., Ganesan, K. A., et al. “EntityBases: Compiling, Organizing and Querying Massive Entity Repositories.” IC-AI (CSREA Press), 2007.
This paper describes methods for constructing and querying large-scale entity repositories, supporting more efficient search and retrieval.
Question Answering Using Vector-Based Information Retrieval with Word Sense Disambiguation
Kavita A. Ganesan
Citation: Ganesan, K. A. “Question Answering Using Vector-Based Information Retrieval with Word Sense Disambiguation.” USC/ISI, 2006.
This technical report combines vector-based retrieval with word sense disambiguation to improve the accuracy of question answering systems.
Automated Story Capture from Conversational Speech
A. S. Gordon, Kavita A. Ganesan
Citation: Gordon, A. S., Ganesan, K. A. “Automated Story Capture from Conversational Speech.” K-CAP, 2005.
Explores techniques to automatically transform conversational speech into structured narrative summaries.