image

Vanakkam, Welcome! 😊

Hello, thank you for landing here and welcome to my homepage.
-- I am a data scientist with extensive experience in different domains - FinTech, Healthcare, EdTech, E-commerce and Retail.
-- I am an academic as well - I like to research, publish and teach. I currently teach courses related to Data Science and Statistics for various top universities/edtech forums.
-- My other interests include Music, Cricket and Reading. I play the Guitar and the Piano; Chennai Super Kings fan in the IPL and Yuval Noah Harari is my favorite author!
-- Take a ride down this page to know more about me.

Current Position

  • Manager, Machine Learning

    PayPal
    Chennai, India

    Jan 2021 - Current

  • Visiting Faculty

    IIM Calcutta

    Batch: MBA-Ex 2022, MBA-Ex 2023
    Course: Business Data Mining

    IIT Madras

    Batch: Online Bsc in Data Science
    Courses: Business Data for Management, Business Analytics, Tools in Data Science, FinTech

    IIM Tiruchirappalli

    Batch: PGCBAA-01
    Course: Python programming for AI and ML

    IIM Kozhikode

    Batch: GMIT 2022;
    Course: AI for Business

    IIM Raipur

    Batch: ePGP 2022 (Executive MBA)
    Course: Machine Learning for Managers

    XLRI

    Batch: PGCBA - 3,4,5;
    Courses: Data Mining, Machine Learning

    • Work Experience

  • Senior Data Scientist

    ZS Associates
    Bengaluru, India

    Jun 2019 - Jan 2021

  • Research Scholar/Teaching Assistant

    Indian Institute of Technology Madras (IIT Madras)
    Chennai, India

    Aug 2014- Jun 2019

  • Applied Scientist Intern

    Amazon
    Bengaluru, India

    May 2018 - Aug 2018

  • Research Intern

    The Southern India Chamber of Commerce and Industry (SICCI)
    Chennai, India

    May 2016 - Jul 2016

  • Systems Engineer

    Tata Consultancy Services Ltd. (TCS)
    Chennai, India

    Aug 2012 - Aug 2014

Education

  • Ph.D.

    Management Information Systems,

    Indian Institute of Technology Madras (IIT Madras)

  • M.S. (By Research)

    Management Information Systems,

    Indian Institute of Technology Madras (IIT Madras)

  • B.E.

    Electrical and Electronics Engineering,

    Anna University

Research Summary

Some of my recent publications in reputed journals and conferences in Information Systems include Information Systems Frontiers (Springer; ABDC A), Transactions on Human-Computer Interaction (AIS, ABDC A) and Data Base for Advances in Information Systems (ACM, ABDC A). I have presented my research work in several elite research forums across the globe and won some prestigious awards and scholarships for outstanding research including Overall Best Paper Award in 22nd Americas Conference on Information Systems (AMCIS 2016 – San Diego, US) and HICSS Doctoral Scholar Award in 51st Hawaii International Conference on System Sciences (HICSS 2018 – Hawaii, US). My PhD dissertation was also nominated for the ICIS SIGMIS Best Doctoral Dissertation Award (2019 – Munich, Germany).

I am also an expert reviewer for reputed journals and conferences including European Journal of Information Systems (EJIS), Information Systems Frontiers (ISF), IIM Kozhikode Society & Management Review (IIMK SMR), ICIS (2019,2020), ECIS (2020) and HICSS (2017-2021). My research interests majorly focus on E-commerce, Web Personalization, Healthcare Analytics and Applied Machine Learning.

Interests

  • Web Personalization
  • E-commerce
  • Healthcare Analytics
  • Applied Machine Learning
  • Behavioral Experiments

Academic Publications

Filter by type:

Sort by year:

“DeepGRASS: Graph, Sequence and Scaled Embeddings on large scale transactions data”

Mahesh Balan U, Vignesh Ravichandran, M. Rohith Srinivaas, Venkat Subramanian Selvaraj
Conference Papers 2021 Swedish Workshop on Data Science (SweDS), 2021, pp. 1-6,. IEEE.

Abstract

Representation learning has redefined large scale data mining applications. The high dimensional embeddings learn complex associations that transcend the human cognitive understanding and have achieved great success in different business applications that encounter the curse of dimensionality, including fin-tech. Different algorithms learn embeddings that capture different types of associations, and it would be useful to learn embeddings that holistically learn multi-dimensional associations. In this paper, we propose DeepGRASS – an algorithm that embeds financial transactions using graph and sequence-based topologies. Our results show that these embeddings learn associations that are very comprehensive, holistic, and multi-dimensional.We deploy DeepGRASS in PayPal, and train it on multitude of transaction data with multi-dimensional features. The algorithm is two-fold: it embeds a bipartite graph with customer and merchant nodes and parallelly learns sequential associations using historical transactions along with other transactional features. These embeddings are then scaled and combined to learn multidimensional associations. We tested this on different predictive applications and find that the learning is generic and shows benchmarking performance in different predictive contexts. Based on offline metrics, back-tests, and sensitivity analysis on offline transaction data, we find very strong evidence to suggest that these embeddings provide the highest AUC score in predictive applications, highest co-efficient of determination in explaining variance and the features explain different types of associations. To our knowledge, this is the first application of embeddings that learn both graph and sequence-based associations on large scale financial transaction data and paves the way for a new generation of feature engineering in fin-tech.

“The Persuasive Nature of Web Personalization on Online Users’ Product Perception: A Mental Accounting Perspective”

Mahesh Balan U, Saji K Mathew
Journal Papers Transactions on Human-Computer Interaction, AIS.

Abstract

E-commerce firms strive to enhance engagement by providing augmented experiences to online users. This research focuses on one such shopping experience enhancement technique—Web personalization. In this study, we examine how personalization affects online users’ perceptions and how different personalization levels differentially impact those perceptions. Drawing on mental accounting theory, we argue that personalization, by providing convenience in online buying, increases transaction utility and, thus, influence online users’ product perceptions. We conducted a laboratory experiment in a public university in Southern India where users took buying decisions at four different personalization levels: zero, low, medium, and high. The findings from this study suggest that product prices affect users’ perceived product quality, which, in turn, affects their perceived product values and, subsequently, their final purchase decision. Web personalization plays a moderating role in all cause-effect relations above. This study contributes to the existing literature on the Web personalization strategy and online user behavior. We find empirical evidence to show that personalization plays a moderating role in the relationship between user perception and intention to purchase.

“Predicting Unplanned Hospital Readmissions using Patient Level Data”

Mahesh Balan U, Meet Gandhi, Swaminathan Rammohan
Conference Papers Proceedings of the Proceedings of the 54th Hawaii International Conference on System Sciences (pp. 3436-3444). AIS.

Abstract

The rate of unplanned hospital readmissions in the US is likely to face a steady rise after 2020. Hence, this issue has received considerable critical attention with the policy makers. Majority of hospitals in the US pay millions of dollars as penalty for readmitting patients within 30 days due to strict norms imposed by the Hospital Readmission Reduction Program. In this study, we develop two novel models: PURE (Predicting Unplanned Readmissions using Embeddings) and Hybrid DeepR, which uses the historical medical events of patients to predict readmissions within 30 days. Both these models are hybrid sequence models that leverage both sequential events (history of events) and static features (like gender, blood pressure) of the patients to mine patterns in the data. Our results are promising, and they benchmark previous results in predicting hospital readmissions. The contributions of this study add to existing literature on healthcare analytics.

“Personalize, Summarize or Let them Read? A Study on Online Word of Mouth Strategies and Consumer Decision Process”

Mahesh Balan U, Saji K Mathew
Journal PapersInformation Systems Frontiers, Springer.

Abstract

This study integrates theories of task complexity and cognitive stopping rule to understand how complexity of information environment impacts uncertainty, effectiveness and efficiency of consumers’ decision process. Using the reviews provided by an online retailer, we develop an e-commerce environment with three levels of complexity: high with raw textual reviews, medium with attribute-level review summaries and low with web personalization strategy based on attribute preferences extracted from online reviews. In a controlled lab experiment, users took buying decisions under different levels of complexity. Our analyses of clickstream data showed that users’ effectiveness and efficiency were the highest in review based personalized environment. However, between groups who received summarized and textual reviews, the latter demonstrated apparently higher effectiveness and efficiency in decision making, which went against our anticipation. Further investigation showed that users simplified decision process when exposed to raw reviews. These results further inform reviews-based personalization strategy in e-commerce.

“Is Cash Still the Enemy? The Dampening of Demonetization’s Ripple Effect on Mobile Payments”

Mahesh Balan U, Abhipsa Pal
Conference Papers Proceedings of the International Working Conference on Transfer and Diffusion of IT (pp. 529-540). Springer.

Abstract

Technology diffusion has often been triggered unintendedly by crises and disasters, as witnessed in several cases including the demonetization cashcrisis surging mobile payment adoption in India. However, once the shock waves induced by the crisis event weakens over time, there exists a void that questions the sustenance of the technology whose diffusion was a ripple effect of the shock. This is seldom explored by the literature that focuses on the immediate aftermath of the crises. We address this limitation by examining the cash withdrawal patterns from ATMs in India post-demonetization for a continuous period of three years. The results provide strong empirical evidence to support our claims towards the dampening of demonetizations’ ripple effect on mobile payments. The theoretical contributions of the study add further to the existing literature on technology diffusion and technology adoption post-crises with a focus on the digital payment systems. The findings have implications for policymakers and government concerned with the digital economy, with cash emerging as an enemy overshadowing the growth of digital payment methods.

“An Experimental Study on the Swaying Effect of Web-Personalization”

Mahesh Balan U, Saji K Mathew
Journal PapersACM SIGMIS Database: the DATABASE for Advances in Information Systems, ACM.

Abstract

The multi-dimensionality of online word-of-mouth not only provides rich attribute-level information but also influences the attribute preference construction of the online consumer. Though prior research affirms that consumer reviews impact the attribute preference assessment of a consumer in a non-personalized single-product environment, in a personalized, multiple alternative environment, consumers' behavior could be completely different and requires separate attention. Building on the information processing approach and constructive preference perspective, our research analyzes how personalization influences this swaying effect, i.e., the influence of personalization on the attribute preference of a consumer. We conducted a multi-group experiment with four different types of personalization - non-personalized information (no personalization), self-referent information, relevant information, and both self-referent and relevant information. Our results show evidence of a swaying effect of personalization on consumers' attribute preference for products. We found that users, when exposed to different types of personalization, experience different levels of the swaying effect on their attribute preferences of the product. This study contributes significantly to the current discourse on the setbacks of web personalization and also informs practicing managers on how to develop recommender system strategies.

“Impact of Product Review Summarization and Personalization on Online Consumers’ Decisions: An Experimental Study”

Mahesh Balan U, Saji K Mathew
Conference Papers Proceedings of the 2017 ACM SIGMIS Conference on Computers and People Research - SIGMIS-CPR '17.

Abstract

E-commerce platforms have a plethora of information available online in the form of online product reviews and product ratings. Although these product reviews are very helpful to consumers for making their buying decisions, it becomes very difficult to extract relevant information pertaining to product features. In this study, we extend personalization to online product reviews by integrating literature from information systems, computer science and social psychology to understand how task complexity varies in different information environments. In a controlled laboratory experiment users took buying decisions in three different information environments: unstructured voluminous reviews, structured with aspect level summarization of reviews and personalized information by providing personalization based on the stated preferences of the user. Our analyses showed that the cognitive efforts of the users and hence, the complexity of the buying task were significantly related to the information presentation format. The personalized information environment had the lowest cognitive effort. However, the group that had structured reviews in the form of a review summarization had higher cognitive efforts than the group which had unstructured raw textual reviews, which went against our hypothesis. Our focus group discussion provided an explanation to this behaviour by observing that users when exposed to information overload simplify the decision process by adopting some heuristic to make their buying decision due to the limited information processing capacity of the working memory. This study contributes to information systems literature by proposing a product review based personalization strategy in e-commerce.

“Impact of personalized review summaries on buying decisions: An experimental study”

Mahesh Balan U, Saji K Mathew
Conference Papers Proceedings of the 22nd Americas Conference on Information Systems (AMCIS), 2016.

Abstract

This study evaluates the impact of personalization of review summaries on consumers’ cognitive efforts and buying decision. Following an experimental procedure we tested four hypotheses pertaining to online buyers’ decision process. Our results show that personalized review summary significantly reduces the information processing effort and information requirements of those who received personalized review summaries as compared to those who did not. This study thus contributes to e-commerce literature on online buyer behavior and recommender systems strategy.

“Impact of web personalization of online word of mouth on buyers' decision process: An experimental study”

Mahesh Balan U, Saji K Mathew
Conference Papers Proceedings of the IEEE International Conference on Management of Innovation and Technology (ICMIT), 2016

Abstract

This study evaluates the impact of personalization of review summaries on consumer's buying decision and cognitive efforts. Following a quasi-experimental procedure we tested two hypotheses pertaining to online buyers' decision process. Our results show that personalized review summary significantly reduces the information processing effort and information requirements of those who received personalized review summaries as compared to those who did not. This study thus contributes to e-commerce literature on online buyer behavior and recommender systems strategy.

“Online word of mouth using text mining: A review of literature and future directions”

Mahesh Balan U, Saji K Mathew
Conference Papers Proceedings of the IEEE Workshop on Computational Intelligence (IEEE WCI), 2015

Abstract

Online word-of-mouth (Online WOM) is extensively growing today, generating a huge corpus of verbal data and numerical product ratings. In this paper we do a detailed review of research on text-mining and online WOM. Since research in this area is still evolving, we add value to this literature by analyzing most of the literature available in this area and identified four different categories of research focus. We also identified various research gaps that exist in each category and hence give directions for future research in text-mining and online WOM.

Consulting/Freelance Projects

Please click on the organization to see more details.

  • Amgen Inc
    image
    Mahesh has led several data science initiatives to improve Sales and Marketing for Amgen Inc through ZS.
  • Crayon Data Pvt Ltd
    image
    Design, development and deployment of a restaurant recommender system.
  • Airloyal Pvt Ltd
    image
    Data science strategy and solutioning for a very popular android Mobile App.
  • Alstom India Pvt Ltd
    image
    Shop floor scheduling and reporting application.
  • Nihon Technology
    image
    Design, development and deployment of course recommender system for their online learning platform learn-chain.com
  • Hand-in-Hand in India
    image
    Social Return on Investment (SROI) analysis for initiatives led by Hand-in-Hand India (NGO) through multiple field studies and data analyses. This study was coordinated by faculty of IIT Madras.
  • ATM Cash Load Forecast
    image
    Forecasted ATM withdrawals to facilitate ATM cash-load optimization in various parts of the country for a very large financial service provider.

Teaching/Training Assignments;
Invited Talks

Please click on the organization to see more details.

  • IIM Calcutta (Teaching)
    image
    Courses: Business Data Mining
    Batch: MBA-Ex 2023, MBA-Ex 2022 (Executive MBA)
    Co-ordinated by Prof. Priya Seetharaman and Prof. Soumyakanti Chakraborty, IIM Calcutta.
  • IIM Kozhikode (Teaching)
    image
    Courses: AI for Business
    Batch: GMIT, 2022
    Co-ordinated by Prof. Vidushi Pandey, IIM Kozhikode.
  • IIM Raipur (Teaching)
    image
    Courses: Machine Learning for Managers
    Batch: ePGP 2022 (Executive MBA)
    Co-ordinated by Prof. Sourya Joyee De, IIM Raipur.
  • IIT Madras (Teaching)
    image
    Instructor for Online B.Sc. in Data Science degree program. Courses: Business Data Management, Business Analytics, Tools for Data Science
  • IIM Tiruchirappalli (Teaching)
    image
    Courses: Python for AI and ML
    Batch: PGCBAA (Executive Program)
    Co-ordinated by Prof. Gajanand, IIM Tiruchirappalli.
  • XLRI Jamshedpur (Teaching)
    image
    Courses: Data Mining, Machine Learning, AI for Digital Transformation
    Batch: PGCBA (Executive Program)
    Co-ordinated by Prof. Ajith and Prof. Giridhar, XLRI.
  • UpGrad
    image
    Offered a course on "Natural Language Processing (NLP)" in UpGrad (AMLC program).
  • IIM Calcutta - Invited Seminar
    image
    Invited Research Seminar, Nov 2020.
    Co-ordinated by Prof. Uttam Kumar Sarkar, IIM Calcutta.
  • IIM Bangalore - Invited Seminar
    image
    Invited Research Seminar, Jan 2020.
    Co-ordinated by Prof. Rahul De, IIM Bangalore.
  • IIM Kozhikode - Invited Seminar
    image
    Invited Research Seminar, Nov 2019.
    Co-ordinated by Prof. Satish Krishnan and Prof. Aravind Sampath.
  • Imarticus Learning (Teaching)
    image
    Machine Learning, Deep Learning, Statistics, R and Python;
    Audience: Graduates, Working executives.
  • NIT Trichy (Training; Workshop)
    image
    Data Mining and Web Mining;
    Audience: Graduate students.
  • Computer Society of India (CSI) (Training; Workshop)
    image
    Digital Marketing;
    Audience: Working executives.

Teaching Assistantships (TA), Coursework and Certifications

  • Teaching Assistantship (TA)

    Data Warehousing and Data Mining (Jul-Sep'15, Jul-Sep'16)

    Management Information Systems (Sep-Nov'15, Sep-Nov'16)

    Information Technology Lab(Sep-Nov'15, Sep-Nov'16)

    Spreadsheet for Business Data Analysis (Jun-Jul'16)

  • Certifications

    Fundamentals of Reinforcement Learning - University of Alberta (2020)

    Deep Learning Specialization - deeplearning.ai (2018)

    Machine Learning - Stanford University (2016)

    Advanced Quantitative Methods for Research - Indian Institute of Management Ahmedabad (2015)

  • Masters and Doctoral Coursework



    Research Courses

    Data Analysis for Research

    Multivariate Statistics for Social Sciences Research

    Research Methodologies in IT

    Decision Theory

    Natural Language Processing



    General Courses

    Data Warehousing and Data Mining

    Management Information Systems

    Microeconomics

    Strategic Management

    Marketing Management

    Financial Management

    Operations Management

    Operations Research

    Heuristics for Decision Making

    French I

Contact Details

  •    Email: maheshbalan[at]alumni[dot]iitm[dot]ac[dot]in