This study examines technology effectiveness for industry demand in which artificial intelligence (AI) is applied in the financial sector. It summarizes prior studies on chatbot and customer service and investigates theories on acceptance attitudes for innovative technologies. By setting variables, the study examines bank revenue methodologically and assesses the impact of customer service and chatbot on bank revenues through customer age classification. The results indicate that new product-oriented funds or housing subscription savings are more suitable for purchase through customer service than through chatbot. However, services for existing products through chatbot positively affect banks’ net income. When classified by age, purchases by the majority age group in the channel positively affect bank profits. Finally, there is a tendency to process small banking transactions through the chatbot system, which saves transaction and management costs, positively affecting profits. Through empirical analysis, we first examine the effect of an AI-based chatbot system implemented to strengthen financial soundness and suggest policy alternatives. Second, we use banking data to increase the study’s real-life applicability and prove that problems in customer service can be solved through a chatbot system. Finally, we investigate how resistance to technology can be reduced and efficiently accommodated.

https://www.mdpi.com/2071-1050/13/6/3173/pdf

Toward a Chatbot for Financial Sustainability

A chatbot is emerged as an effective tool to address the user queries in automated, most appropriate and accurate way. Depending upon the complexity of the subject domain, researchers are employing variety of soft-computing techniques to make the chatbot user-friendly. It is observed that chatbots have flooded the globe with wide range of services including ordering foods, suggesting products, advising for insurance policies, providing customer support, giving financial assistance, schedule meetings etc. However, public administration based services wherein chatbot intervention influence the most, is not explored yet. This paper discuses about artificial intelligence based chatbots including their applications, challenges, architecture and models. It also talks about evolution of chatbots starting from Turing Test and Rule-based chatbots to advanced Artificial Intelligence based Chatbots (AI-Chatbots). AI-Chatbots are providing much kind of services, which this paper outlines into two main aspects including customer based services and public administration based services. The purpose of this survey is to understand and explore the possibility of customer & public administration services based chatbot. The survey demonstrates that there exist an immense potential in the AI assisted chatbot system for providing customer services and providing better governance in public administration services. 

https://link.springer.com/content/pdf/10.1007/s11042-021-11458-y.pdf

A survey on providing customer and public administration based services using AI: chatbot

/pdf/a-survey-on-providing-customer-and-public-administration-1nrob3wk.pdf

Technology has turned into a significant differentiator in the money and traditional recordkeeping systems for the financial industry. To depict two customers as potential investors, it is mandatory to give the complex innovation that they anticipate and urge to purchase. In any case, it is difficult to keep on top of and be a specialist in each of the new advancements that are accessible. By reappropriating IT administrations, monetary administrations firms can acquire prompt admittance to the most recent ability and direction. Financial systems, along with machine learning (ML) algorithms, are vital for critical concerns like secure financial transactions and automated trading. These are the key to the provision of financial decisions for investors and stakeholders for the firms which are working with the trade credit (TC) approach, in Small and Medium Industries (SMEs). Huge and very sensitive data is processed in a limited time. The trade credit is a reason for more financial gains. The impact of TC with predictive machine learning algorithms is the reason why intelligent and safe revenue generation is the main target of the proposed study. That is, the combination of financial data and technology (FinTech) domains is a potential reason for sales growth and ultimately more profit.

https://downloads.hindawi.com/journals/scn/2022/7149902.pdf

Machine Learning: The Backbone of Intelligent Trade Credit-Based Systems

We develop a chatbot using Deep Bidirectional Transformer models (BERT) to handle client questions in financial investment customer service. The bot can recognize 381 intents, and decides when to say \textit{I don't know} and escalates irrelevant/uncertain questions to human operators. Our main novel contribution is the discussion about uncertainty measure for BERT, where three different approaches are systematically compared on real problems. We investigated two uncertainty metrics, information entropy and variance of dropout sampling in BERT, followed by mixed-integer programming to optimize decision thresholds. Another novel contribution is the usage of BERT as a language model in automatic spelling correction. Inputs with accidental spelling errors can significantly decrease intent classification performance. The proposed approach combines probabilities from masked language model and word edit distances to find the best corrections for misspelled words. The chatbot and the entire conversational AI system are developed using open-source tools, and deployed within our company's intranet. The proposed approach can be useful for industries seeking similar in-house solutions in their specific business domains. We share all our code and a sample chatbot built on a public dataset on Github.

/pdf/ava-a-financial-service-chatbot-based-on-deep-bidirectional-1hdrol1lt1.pdf

AVA: A Financial Service Chatbot Based on Deep Bidirectional Transformers

We develop a chatbot using Deep Bidirectional Transformer models (BERT) to handle client questions in financial investment customer service. The bot can recognize 381 intents, and decides when to say "I don't know" and escalates irrelevant/uncertain questions to human operators. Our main novel contribution is the discussion about uncertainty measure for BERT, where three different approaches are systematically compared on real problems. We investigated two uncertainty metrics, information entropy and variance of dropout sampling in BERT, followed by mixed-integer programming to optimize decision thresholds. Another novel contribution is the usage of BERT as a language model in automatic spelling correction. Inputs with accidental spelling errors can significantly decrease intent classification performance. The proposed approach combines probabilities from masked language model and word edit distances to find the best corrections for misspelled words. The chatbot and the entire conversational AI system are developed using open-source tools, and deployed within our company's intranet. The proposed approach can be useful for industries seeking similar in-house solutions in their specific business domains. We share all our code and a sample chatbot built on a public dataset on Github.

A Financial Service Chatbot based on Deep Bidirectional Transformers

The fundamental principle in Modern Portfolio Theory (MPT) is based on the quantification of the portfolio's risk related to performance. Although MPT has made huge impacts on the investment world and prompted the success and prevalence of passive investing, it still has shortcomings in real-world applications. One of the main challenges is that the level of risk an investor can endure, known as \emph{risk-preference}, is a subjective choice that is tightly related to psychology and behavioral science in decision making. This paper presents a novel approach of measuring risk preference from existing portfolios using inverse optimization on the mean-variance portfolio allocation framework. Our approach allows the learner to continuously estimate real-time risk preferences using concurrent observed portfolios and market price data. We demonstrate our methods on real market data that consists of 20 years of asset pricing and 10 years of mutual fund portfolio holdings. Moreover, the quantified risk preference parameters are validated with two well-known risk measurements currently applied in the field. The proposed methods could lead to practical and fruitful innovations in automated/personalized portfolio management, such as Robo-advising, to augment financial advisors' decision intelligence in a long-term investment horizon.

Learning Time Varying Risk Preferences from Investment Portfolios using Inverse Optimization with Applications on Mutual Funds.

Low-rank inductive matrix completion (IMC) is currently widely used in IoT data completion, recommendation systems, and so on, as the side information in IMC has demonstrated great potential in reducing sample complexity and computational complexity. Unfortunately, the saddle point remains a major obstacle for the convergence of the nonconvex solutions to IMC. What’s more, carefully choosing the initial solution alone does not usually help remove the saddle points. To address this problem, we propose a stochastic variance reduction gradient-based algorithm called LRSVRGIMC. LRSVRG-IMC can escape from the saddle points under various low-rank and sparse conditions with a properly chosen initial input. We also prove that LRSVRG-IMC achieves both a linear convergence rate and a near-optimal sample complexity. The superiority and applicability of LRSVRG-IMC are verified via experiments on synthetic datasets.

LRSVRG-IMC: An SVRG-Based Algorithm for LowRank Inductive Matrix Completion

Forecasting stock prices has been a prominent research topic in finance for many years. With the rapid advancement of internet technologies, financial news has become a crucial source of information for investors, and sentiment analysis of news articles has shown considerable potential in predicting future stock price movements. In this study, we developed a model that combines a Naive Bayes sentiment classifier to convert news text into sentiment indicators, and an LSTM neural network model to predict stock price movements. We obtained data from Oriental Fortune‘s financial commentary bar, covering the period from 28 July to 23 December 2022, resulting in a total of 318,373 data points. Then, we selected 1400 news items and manually labeled them to train a Naive Bayes classifier to quantify the news sentiment of each trading day. Our feature variables include sentiment indicator, closing price, price movements,  ratio, and  ratio of SSE 50 constituents. Finally, we applied the Long and Short-Term Memory (LSTM) model to predict stock price movements and compare their performance with the ARIMA time series method. Our results demonstrate that the LSTM model outperforms the traditional ARIMA model in predicting stock price movements. Furthermore, we compared the LSTM model with and without sentiment indicators and found that the inclusion of sentiment indicators significantly improves forecasting its performance. Overall, our proposed model offers a promising approach to predicting stock movements and provides insights into the effectiveness of incorporating sentiment analysis in financial forecasting.

Yuxin Chen

Papers

A Financial Service Chatbot based on Deep Bidirectional Transformers

AVA: A Financial Service Chatbot Based on Deep Bidirectional Transformers

Learning Time Varying Risk Preferences from Investment Portfolios using Inverse Optimization with Applications on Mutual Funds.

LRSVRG-IMC: An SVRG-Based Algorithm for LowRank Inductive Matrix Completion

LSTM-based Stock Price Prediction Model using News Sentiments