Technology innovations meet greatest success in business when these are entirely ‘client focussed’. Developments in the retail sector, which is consumer-led, are addressing client demand for more personalised, faster and competitive services. Artificial Intelligence, Machine Learning and Sentiment Analysis are changing the way in which these services are offered. In particular Financial Organisations are creating and leveraging such innovation in the domain of wealth management. This trend is now being taken on board by multiple innovators: academia, start-ups, technology companies and financial market participants.
Sentiment Analysis (SA) is a technology that makes rapid assessment of the sentiments expressed in news releases as well as other media sources such as Twitter and blogs. This conference addresses and explains how to extract sentiment from the multiple sources of information and showcases the advances that have taken place in the field of financial innovation, as well as new applications in the domain of Retail and Consumer Marketing.
| Day 1 | Day 2 | |
|---|---|---|
| Consumer Markets Stream | Consumer Markets – Research & Academic Focus [Cnm – Res] |
Consumer Markets – Industry Focus [Cnm – Ind] |
| Financial Markets Stream | Financial Markets – Industry Focus [Fin – Ind] |
Financial Markets – Research & Academic Focus [Fin – Res] |
The conference is being organised in collaboration with partner organisations that have considerable experience in the field and bring new insights and interested parties to the table. In Bangalore, the partner is the well-respected Indian Institute of Management (IIM) Calcutta which is home to the Financial Research & Trading Lab, Indian Institute of Management (IIM) Bangalore.
After the great success of the London and Singapore events, we are now bringing the conference to Bangalore[Bengaluru]. Bangalore[Bengaluru] is often referred to as the “Silicon Valley of India” (or “IT capital of India”) because of its role as the nation’s leading IT exporter. So this makes it a highly appropriate venue with access to a large pool of qualified and enthusiastic professionals.
Ajit Balakrishnan, Chief Executive Officer, Rediff.com
What can R and Big Data do to throw new light on classic marketing challenges such as online consumer market segmentation and product recommendations... I will present a few cases of such applications using large scale Indian data and also cast an eye on where Deep Learning could go next.
Nitish Sinha, Senior Economist, Federal Reserve Board USA
In this talk I will discuss a co-authored work with Steve Heston that uses a dataset of more than 900,000 news stories to test whether news can predict stock returns. We measure sentiment with a proprietary Thomson-Reuters neural network. We find that daily news predicts stock returns for only 1 to 2 days, confirming previous research. Weekly news, however, predicts stock returns for one quarter. Positive news stories increase stock returns quickly, but negative stories have a long-delayed reaction. Much of the delayed response to news occurs around the subsequent earnings announcement. The analysis and conclusions set forth are those of the speaker and do not indicate concurrence by other members of the research staff or the Board of Governors.
Xiang Yu, Business Development Techno Executive, and Gautam Mitra, CEO/Director, OptiRisk Systems/UCL, UK
We have created an innovative and dynamic trading strategy for equities, with a particular focus on controlling downside risk. The mathematical concept behind the approach is called stochastic dominance, where investment decisions are based on distributions rather than moments. A major contribution of news sentiment is in the prediction of future distributions. Regression analysis on news sentiment and regime switching models are employed to digest market moods and account for changing market situations.
Richard Peterson, CEO, MarketPsych Data, USA
Dr. Peterson will describe the unique characteristics of media sentiment data and approaches to financial price prediction with this data. The basics of media sentiment data, various modeling approaches, and their results (including live trading results) will be described in this talk. Viewers will gain an understanding of real-world modeling tips and techniques when dealing with noisy and inconsistent data such as media sentiment streams.
Ashok Banerjee, Departmental Head of Finance and Control, Indian Institute of Management, Calcutta (IIMC), India
Banks and financial institutions in emerging markets are saddled with a huge proportion of bad loans. Banking regulations require lenders to provide for troubled debt which adversely affects the profitability of banks. The capital market also reacts negatively to such write-offs of big ticket debts. Banks are, therefore, putting significant resources into developing early warning signals to arrest eventual default. The financial institutions use a wide range of default prediction models to estimate the loan loss. These models use data from financial statements and the market. The present study shows that such models fail to provide effective early warning signals. We use annual reports of companies to develop a default model which is predictive and hence has the capability of providing early warning signals. Using information from Directors' Reports, Audit Reports and notes to accounts, our model successfully discriminates the 'good' firms from the 'bad' ones.
Enza Messina, Professor in Operations Research, University of Milano-Bicocca, Italy
We show how deep learning and ensemble methods can successfully address challenging problems arising in sentiment analysis such as irony detection or domain adaptation. In particular, we propose an unsupervised framework for domain-independent irony detection built upon an existing probabilistic topic model initially introduced for sentiment analysis purposes. Moreover, in order to improve its generalization abilities, we apply Word Embeddings to obtain domain-aware ironic orientation of words. The acquisition of cross-domain high level feature representations through word embeddings combined with the generalization capability of ensemble methods can also be used for addressing the problem of domain adaptation also in the scenario where the testing target domain is completely unlabeled.
Deepak Mishra, Head of Customer Analytics & Platform - Asia, Thomson Reuters
In general, attributes related to customer behaviour with organizations are primarily used to develop predictive models for retention and growth. Macroeconomic factors are also utilized to enhance the predictive power of the model, but these factors provide “blanket” attributes, the same for all customers in a country, sector etc. Using more information from social media, news, mergers and acquisitions, IPO etc. can enhance the predictive power of the model.
Praloy Mujumder, CEO, Disseminare Consulting
Prateek Agrawal and Surangama Agarwal, IVY Professional School
Recommender system is a powerful method to extract and filter through large information and various algorithms has been built to help us understand the process to embed recommender technology in specific domains. Collaborative Filtering is one such concept which led to building of various algorithms. It bases its predictions and recommendations on the ratings or behavior of other users in the system and helps to provide personalized recommendations.
Vivek Bajaj, Director, Kredent Ventures
Financial markets offer tremendous data points that needs regular coding & decoding. The whole industry thrives on the probability of predicting the future based on the past data points. In today's fast moving world with shorter attention time span, a participant needs a strong tool to comprehend data faster and effectively. Stockedge is doing something interesting for Indian equity market participants. With over 150000+ downloads in just 6 months and 4.7 rating by over 3000 users, the tool is empowering Indian investors in true sense.
Niraj S Kakkad, AVP – Private Wealth Advisory, InvestAscent Wealth Advisors Pvt Ltd
Artificial intelligence (AI) is thinking capabilities built into machines which help in logical reasoning and problem solving. AI is an imminent change in the Indian financial markets, advisors must embrace it. AI can help comprehend a client’s expectations from investments by helping with cognitive deductions of risk appetite, deriving portfolio weightage, fund selection etc. I intend to present a Critical SWOT analysis of the impact of AI on Indian Financial Markets - focused on client advisor relationship.
Moderator: Gautam Mitra
Ashok Banerjee, Departmental Head of Finance and Control, Indian Institute of Management, Calcutta (IIMC), India
Studies at the Finance Lab of IIM Calcutta show that the effect of any news on the market depends on the attention of investors. If the attention of investors were somewhere else, even news carrying strong positive/negative sentiment would go unnoticed or would be penalized. This is particularly true in the case of major non-market attention grabbing events. In other words, attention overwhelms the effect of sentiment. Processing any attention-grabbing event requires effort. If that effort is directed towards some particular information, people are too lazy to put inthe extra effort to process any other information at the same time, no matter how much sentiment that information might carry.
Arup Ganguly, PhD Candidate, Katz Graduate School of Business, University of Pittsburgh, USA
Prior theoretical and empirical studies in both finance and accounting are split on the relation between information disclosure and litigation risk. I argue that more information is disclosed in the form of textual data (non-numerical form) through SEC filings that has not been analyzed in most of the extant studies. Using hand-collected data on federal civil securities class action lawsuits for the period 1996-2014 and widely used techniques in natural language processing and propensity score matched sample to address endogeneity concerns to some extent, I study whether the degree of information disclosure through texts in SEC filings (10-Ks and 10-Qs), the readability of such disclosures, and the sentiments generated through the choice of words, are associated with the incidence of securities class action lawsuits. I find results that are consistent with the theoretical view that argues that greater disclosure is often perceived as ex post misleading, precipitating securities class action litigations. Moreover, I find that despite the boiler plate nature of SEC filings, both readability of the text used and sentiments generated through the selection of words have a significant predictive power in explaining the likelihood of being sued by shareholders in class actions. Lastly, I document how managers alter their behavior with respect to textual disclosures pre- versus post-litigation using a standard difference-in-differences (DiD) framework. These results are robust to the use of different empirical specifications, controls and various measures of textual disclosure, readability and sentiments.
Analytics and Intuition - substitutes or complements in decision making?
Panel Members:1. Kingshuk Banerjee, IBM Global Business Services;
2. Pavinder Monga, Citi;
3. Anup Gunaseelan, LatentView Analytics;
4. Gaurav Singh, Verloop
5. Moderator: Prithwiraj Mukherjee
Svetlana Borovkova, Associate Professor of Quantitative Finance, VrijeUniversiteit Amsterdam, Netherlands
This talk shows how we use media sentiment to measure risk in the global financial system. I introduce a new measure of systemic risk called SenSR (for Sentiment-based Systemic Risk) and demonstrate that this measure gives an early warning about financial system distress. I then discuss perceived financial networks, which we build using media attention directed to banks, and whose characteristics can also help us understand systemic risk. Finally, I address the construction of sentiment-based indicators, similar toSenSR, for other industries and their use as new factors for investment purposes.
M. Jeevananthan and Monika MS, Thiagarajar School of Management
Keshav Sehgal, Walmart Data Science Labs, Bangalore
Deep neural networks in the recent years have successfully solved challenging problems from various domains. Prediction of stock price movements in financial markets is one of the challenging problems attempted by researchers. There have been limited attempts to use Deep Learning in the context of stock movement predictions. The current paper probes into using deep learning techniques to predict the daily price movements of equity stocks. Multi-layer perceptron and Recurrent Neural Networks were trained on price data of equities. The paper discussed detailed methodology of training and testing process. Both MLP and RNN model performed weak learners when assessed over test data. The results were consistently above the baseline accuracy.
Sankarshan Basu, Professor in the Finance and Accounting Area at the Indian Institute of Management, Bangalore (IIMB), India
Technology has significantly altered the way the world has known finance over the years. While this has brought significant benefits to the system and the economies as a whole, it has thrown up several challenges as well – some having quite disastrous consequences. This talk will dwell upon the benefits that technology has provided to the financial sector in general and the banking sector in particular at the same time highlighting the pitfalls that have sprung up in the process. Part of the talk also looks at how technology can be used in a dynamic environment context to address some of the issues related to the pitfalls and more particularly what, if any, measures can be taken to reduce the pitfalls in the future.
Ajit Balakrishnan, Chief Executive Officer, Rediff.com
What can R and Big Data do to throw new light on classic marketing challenges such as online consumer market segmentation and product recommendations... I will present a few cases of such applications using large scale Indian data and also cast an eye on where Deep Learning could go next.
Nitish Sinha, Senior Economist, Federal Reserve Board USA
In this talk I will discuss a co-authored work with Steve Heston that uses a dataset of more than 900,000 news stories to test whether news can predict stock returns. We measure sentiment with a proprietary Thomson-Reuters neural network. We find that daily news predicts stock returns for only 1 to 2 days, confirming previous research. Weekly news, however, predicts stock returns for one quarter. Positive news stories increase stock returns quickly, but negative stories have a long-delayed reaction. Much of the delayed response to news occurs around the subsequent earnings announcement. The analysis and conclusions set forth are those of the speaker and do not indicate concurrence by other members of the research staff or the Board of Governors.
Prithwiraj Mukherjee, Indian Institute of Management, Bangalore
Modelling customer churn is important for retailers, especially when dropout is unobserved. We compare two popular methods used by managers – the BG-NBD model (Fader, Hardie and Lee 2005) that uses recency and frequency as inputs, and Markov Chain models incorporating threshold recency as a dropout assumption. We compare these approaches on parameters like accuracy and computational load across multiple data sets.
Avadhoot Jathar, Senior Statistician, Analytics Quotient
Demand faced by organized retailers in emerging markets comes from both households and resellers. Retailer’s price promotions must not be left to leave money on the table. Descriptively, undirected classification highlights uptake by resellers, and we further present this unique context of demand aggregation. Our analysis suggests deals with quantity ceilings can be a useful pricing tool for retailer’s category management.
Kunal Saxena, Associate Professor of Marketing, and Varsha PS, Assistant Professor, Alliance University, Bangalore
This presentation gives a brief outline of the history and applications of Artificial Intelligence (AI) in numerous areas. It presents some real-world examples of artificial intelligence in consumer markets, such as marketing/advertising applications with significant traction and burgeoning future applications for AI in marketing / advertising.
Nishant Chandra, R&D Scientist, AIG
The latest advances in natural language understanding has created a massive paradigm shift in dealing with text related data problems. Deep learning has created a revolution in the NLU space and corporations are leveraging it in various ways. The technology barrier is significantly reduced with open source technologies that are easy to configure and use. Several open source tools are available in the machine learning domain for traditional natural language processing to deep learning. Helpful implementation tips will be provided along with evaluating the technologies and tools.
Arun Mallavarapu, Fedo
Kalya Lakshmi Sainath, Lloyds Business School and Korcha Teja Sai, Sri Venkateswara University
This study supports sentiment analysis as additional research technique for accumulating and analysing recorded data on the internet. Sentiment analysis is a data excavating technique that steadily gauges textual content using machine learning techniques. As a research method in marketing, sentiment analysis provides a proficient and real valuation of consumer views in real time. It allows data collection and analysis from a huge sample without interferences, impediments and time delays.The paper concludes with the challenges marketeers can face when using this practice in their research work.
Madhu Gopinathan, Vice President, Data Science, MakeMyTrip
The algorithms for analyzing sentiment ranges from using simple wordlists with positive and negative words to employing our brains for deep understanding of natural language text. This talk focuses on how we can go beyond mere word level analysis to concept level analysis and then extracting relations between concepts for improving sentiment analysis.
Gaurav Gaba, AVP at Société Générale Global Solution Centre
This presentation covers the Genesis, Conventional Approach and Change to Transformational Risk Management Practices via Artificial Intelligence with machine learning and sentiment analysis capabilities. It further gives a view of the future where the focus lies on pure quality, efficiency and practices built on knowledge which is a foot wide but a mile deep.
Analytics and Intuition - substitutes or complements in decision making?
Panel Members:1. Kingshuk Banerjee, IBM Global Business Services;
2. Pavinder Monga, Citi;
3. Anup Gunaseelan, Manager, LatentView Analytics;
4. Gaurav Singh, Verloop
5. Moderator: Prithwiraj Mukherjee
Robin Panicker, CEO, SEQATO Software Solutions
In this presentation, we discuss how we can use artificial intelligence and predictive analysis to provide accurate insights that will help consumers in optimised decision making. Machines cannot beat the experience and expertise of a human brain. At the same time, humans cannot beat the speed and perseverance of an automated algorithm. Combining the two leads to production of accurate, optimal and best of the breed solutions. We will talk about how some complicated use-cases in financial and healthcare domain can be cracked easily with the help of AI and ML.
Shabbir Tayabali, Senior Manager - Analytics, Oracle
Decision Tree is a proven and favorite tool among decision makers when the decision making process is complex (includes multiple variables, multiple alternatives and multiple dependencies). Today technologies have empowered customers in various ways thereby impacting their purchase process and the very fundamental purchase funnel. On the other hand, technological advancements have also resulted in various sources of information that suppliers can tap into to understand customer better. However, for this, they have to use various ML algorithms and the one they will always need to understand the customer’s mind is the Decision Tree. In collaboration with other machine learning algorithms, Decision Tree can give a deeper view into the minds of customers that other algorithms cannot.
Krishma Singla, Managing Consultant- Data Science and Cognitive computing, IBM Global Business Services
Krishma presents a case study of how text analytics and natural language processing can massage unstructured information and generate quantifiable measures to provide direction to top management of financial institutions to get a clear edge in business. The next gen machine learning and AI platforms are striving to scale the expertise and augment humans by putting cognitive engines at heart of large scale operations such as smart help desks, perceptive dialogues, advisors/recommendation systems etc.
Prateek Agrawal and Surangama Agarwal, IVY Professional School
Recommender system is a powerful method to extract and filter through large information and various algorithms has been built to help us understand the process to embed recommender technology in specific domains. Collaborative Filtering is one such concept which led to building of various algorithms. It bases its predictions and recommendations on the ratings or behavior of other users in the system and helps to provide personalized recommendations.
Suman Singh, Chief Analytics Officer, Zafin
Banks are increasingly under pressure to provide a more personalized and improved customer experience. With eroding revenue streams, intensifying competition, and ever-increasing customer expectations, financial institutions need to explore a new way of doing business. Advanced statistical methods and machine learning algorithms can help banks to become smarter and more diligent in dealing with their customers. Measuring customer relationship, evaluating the customer journey and recommending right bundle of product & services at the right price in real time through technology-enabled digital platforms, will be the vital enablers to improve the customers’ banking experience.
Vishesh Nigam, GM - Global Strategic Assets Area, Concentrix
Autonomous interactive technologies are helping enterprises to solve their Business Problems and transforming an experience into conversational, personalized, and instantly gratifying engagement for customers and workforce. This new technology goes under different names, e.g. Conversational Bots, ChatBot, Virtual Customer Assistance(VCA) and Virtual Interaction etc.
This presentation addresses current requirements, as organisations seek to understand which is the right omni-channel tool to extend concierge-like services; how to create the best customer experience by selecting the most appropriate tool, as well as to best deploy Natural Language Processing, Information retrieval and continuous improvement through artificial intelligence and sentiment analytics.
