|
on Big Data |
| By: | Xiong Xiong; Fan Yang; Li Su |
| Abstract: | Livestreaming commerce, a hybrid of e-commerce and self-media, has expanded the broad spectrum of traditional sales performance determinants. To investigate the factors that contribute to the success of livestreaming commerce, we construct a longitudinal firm-level database with 19, 175 observations, covering an entire livestreaming subsector. By comparing the forecasting accuracy of eight machine learning models, we identify a random forest model that provides the best prediction of gross merchandise volume (GMV). Furthermore, we utilize explainable artificial intelligence to open the black-box of machine learning model, discovering four new facts: 1) variables representing the popularity of livestreaming events are crucial features in predicting GMV. And voice attributes are more important than appearance; 2) popularity is a major determinant of sales for female hosts, while vocal aesthetics is more decisive for their male counterparts; 3) merits and drawbacks of the voice are not equally valued in the livestreaming market; 4) based on changes of comments, page views and likes, sales growth can be divided into three stages. Finally, we innovatively propose a 3D-SHAP diagram that demonstrates the relationship between predicting feature importance, target variable, and its predictors. This diagram identifies bottlenecks for both beginner and top livestreamers, providing insights into ways to optimize their sales performance. |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2310.19200&r=big |
| By: | Ryan Chipwanya |
| Abstract: | The stock market has been established since the 13th century, but in the current epoch of time, it is substantially more practicable to anticipate the stock market than it was at any other point in time due to the tools and data that are available for both traditional and algorithmic trading. There are many different machine learning models that can do time-series forecasting in the context of machine learning. These models can be used to anticipate the future prices of assets and/or the directional bias of assets. In this study, we examine and contrast the effectiveness of three different machine learning algorithms, namely, logistic regression, decision tree, and random forest to forecast the movement of the assets traded on the Japanese stock market. In addition, the models are compared to a feed forward deep neural network, and it is found that all of the models consistently reach above 50% in directional bias forecasting for the stock market. The results of our study contribute to a better understanding of the complexity involved in stock market forecasting and give insight on the possible role that machine learning could play in this context. |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2310.16855&r=big |
| By: | Philippe Goulet Coulombe (University of Quebec in Montreal); Mikael Frenette (University of Quebec in Montreal); Karin Klieber (Oesterreichische Nationalbank) |
| Abstract: | We reinvigorate maximum likelihoode stimation (MLE) for macroeconomic density forecasting through a novel neural network architecture with dedicated mean and variance hemispheres. Our architecture features several key ingredients making MLE work in this context. First, the hemispheres share a common core at the entrance of the network which accommodates for various forms of time variation in the error variance. Second, we introducea volatility emphasis constraint that breaks mean/variance indeterminacy in this class of overparametrized nonlinear models. Third, we conduct a blocked out-of-bag reality check to curb overfitting in both conditional moments.Fourth, the algorithm utilizes standard deep learning software and thus handles large datasets – both computationally and statistically. Ergo, our Hemisphere Neural Network (HNN) provides proactive volatility forecasts based on leading indicators when it can, and reactive volatility based on the magnitude of previous prediction errors when it must. We evaluate point and density forecasts with an extensive out-of-sample experiment and benchmark against a suite of models ranging from classics to more modern machine learning-based offerings. In all cases, HNN fares well by consistently providing accurate mean/variance forecasts for all targets and horizons. Studying the resulting volatility paths reveals its versatility, while probabilistic forecasting evaluation metrics showcase its enviable reliability. Finally, we also demonstrate how this machinery can be merged with other structured deep learning models by revisiting Goulet Coulombe(2022)’s Neural Phillips Curve. |
| Date: | 2023–11 |
| URL: | http://d.repec.org/n?u=RePEc:bbh:wpaper:23-04&r=big |
| By: | Lee, Sangkyu (Korea Institute for Industrial Economics and Trade) |
| Abstract: | When employed in artificial intelligence (AI) applications, machine learning (ML) allows AI to recognize patterns in data and predict future outcomes based on these patterns, supporting decision-making. This additionally allows ML to be utilized in the formulation of industrial policies (IPs). However, overreliance on AI for all pol-icies presents several challenges. To harness AI effectively, it is essential to ensure logical clarity and measurability that can be digitally transformed into data, along with the availability of a sufficient amount of data to ensure accuracy and reliability. On the other hand, it is more difficult to use AI in IP design when policies must take into normative as well as economic considerations, or when it becomes necessary to define new norms. These are typically cases in which simple pattern recognition fails to grasp the complexity of various issues at play, making the immediate application of AI application impossible. For instance, situations in which numerous stakeholders hold diverse perspectives can make it challenging to establish clear policy objectives. Additionally, any given problem may include some issues that are fundamentally subjective or normative, and thus incapably of being quantified or measured. This also presents challenges to the effective use of AI. This paper explores the ways in which machine learning (ML) techniques in the field of object classification can contribute to formulating industrial policies. Thank you for reading this abstract of a report from the Korea Institute for Industrial Economics and Trade! Visit us on YouTube: https://www.youtube.com/watch?v=Q36v30l5CV0 Visit us on Instagram: https://www.instagram.com/worldkiet/ Visit our website: http://www.kiet.re.kr/en |
| Keywords: | artificial intelligence; AI; machine larning (ML); patterns; data; data analysis; pattern recognition; neural networks; industrial policy; policy design; Korea |
| JEL: | E61 E69 I28 L52 L52 L86 L88 |
| Date: | 2023–10–31 |
| URL: | http://d.repec.org/n?u=RePEc:ris:kieter:2023_022&r=big |
| By: | Marcelo DEL Cajias; Anna Freudenreich |
| Abstract: | In this paper, the most influential variables that affect the liquidity (inverse of time on market) of rental apartments are analysed empirically for the city of Munich. Therefore, the random forest machine learning technique based on decision trees is applied. Micro data for more than 100, 000 observations on the residential rental market from 2013 to 2021 is used. As a first step, the main housing, social and spatial predictors of liquidity on the residential rental market are revealed. Results show that the price as well as the size have the greatest impact on the liquidity of residential apartments. From the geographic variables the distances to the next hairdresser, bakery and school are most important. Second, this paper analyses how the survival probability of residential rental apartments responds to these major characteristics. And third, the partial dependency of cost and size on the survival probability is revealed. Hence, the segmentation of dwellings generated by the decision tree methodology results in a deep and profound understanding of the driving factors of liquidity. Although the decision tree methodology has been applied frequently on the real estate market for the analysis of prices, its use for examining liquidity is completely novel. To the best of the authors’ knowledge this is the first paper, to apply a decision tree approach to liquidity analysis on the real estate market. |
| Keywords: | housing; Machine Learning; Random forest; Time on Market |
| JEL: | R3 |
| Date: | 2023–01–01 |
| URL: | http://d.repec.org/n?u=RePEc:arz:wpaper:eres2023_35&r=big |
| By: | Joao Vitor Matos Goncalves; Michel Alexandre; Gilberto Tadeu Lima |
| Abstract: | This paper assesses the impact of time horizon on the relative performance of traditional econometric models and machine learning models in forecasting stock market prices. We employ an extensive daily series of Brazil IBX50 closing prices between 2012 and 2022 to compare the performance of two forecasting models: ARIMA (autoregressive integrated moving average) and LSTM (long short-term memory) models. Our results suggest that the ARIMA model predicts better data points that are closer to the training data, as it loses predictive power as the forecast window increases. We also find that the LSTM model is a more reliable source of prediction when dealing with longer forecast windows, yielding good results in all the windows tested in this paper. |
| Keywords: | Finance; machine learning; deep learning; stock market |
| JEL: | C22 C45 C53 G17 |
| Date: | 2023–11–17 |
| URL: | http://d.repec.org/n?u=RePEc:spa:wpaper:2023wpecon13&r=big |
| By: | Marco Delogu (University of Sassari, IT); Raffaelle Lagravinese (University of Bari, IT); Dimitri Paolini (CRENoS & University of Bari IT, UCL BE); Giuliano Resce (University of Molise, IT) |
| Abstract: | We investigate whether machine learning (ML) methods are valuable tools for predicting students’ likelihood of leaving pursuit of higher education. This paper takes advantage of administrative data covering the entire population of Italian students enrolled in bachelor’s degree courses for the academic year 2013-2014. Our numerical findings suggest that ML algorithms, particularly random forest and gradient boosting machines, are potent predictors pointing to their use as early warning indicators. In addition, feature importance analysis highlights the role of the number of European Credit Transfer System (ECTS) obtained during the first year for predicting the likelihood of dropout. Accordingly, our analysis suggests that policies that aim to boost the number of ECTS gained during the early academic career may be effective in reducing drop-out rates at Italian universities. |
| Keywords: | Early warning system, Machine learning, Dropout; Italy. |
| JEL: | C53 C55 I20 |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:luc:wpaper:22-06&r=big |
| By: | Nian Si |
| Abstract: | In modern recommendation systems, the standard pipeline involves training machine learning models on historical data to predict user behaviors and improve recommendations continuously. However, these data training loops can introduce interference in A/B tests, where data generated by control and treatment algorithms, potentially with different distributions, are combined. To address these challenges, we introduce a novel approach called weighted training. This approach entails training a model to predict the probability of each data point appearing in either the treatment or control data and subsequently applying weighted losses during model training. We demonstrate that this approach achieves the least variance among all estimators without causing shifts in the training distributions. Through simulation studies, we demonstrate the lower bias and variance of our approach compared to other methods. |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2310.17496&r=big |
| By: | Abi Adams-Prassl; Maria Balgova; Matthias Qian; Tom Waters |
| Abstract: | We build a model of job design under monopsony that yields predictions over the relationship between: (i) the amenity value of non-wage job features; (ii) whether they are costly or profitable to firms; (iii) monopsony power. We analyse the amenity value of schedule flexibility offered in the labour market by combining our model’s predictions with a new measure of schedule flexibility, which we construct from job vacancy text using a supervised machine learning approach. We show that the amenity value of schedule flexibility depends crucially on whether it is offered alongside a salaried contract that insures workers from earnings variation. |
| Date: | 2023–02–22 |
| URL: | http://d.repec.org/n?u=RePEc:oxf:wpaper:1002&r=big |
| By: | Jonathan Alexander Muñoz-Martínez; David Orozco; Mario A. Ramos-Veloza |
| Abstract: | This study follows a novel approach proposed by Angelico et al. (2022) using Twitter to measure inflation perception in Colombia in real time. By applying machine learning techniques, we implement two real-time indicators of inflation perception and show that both exhibit a dynamic similar to inflation and inflation expectations for the sample period January 2015 to March 2023. Our interpretation of these results suggests that our indicators are closely linked to the underlying factors that drive inflation perception. Overall, this approach provides a valuable instrument for gauging public sentiment towards inflation and complements the traditional inflation expectations measures used in the inflation–targeting framework. **** RESUMEN: Este estudio sigue un enfoque novedoso propuesto por Angelico et al. (2022) para la medición en tiempo real de la percepción de la inflación en Colombia utilizando Twitter. Mediante la aplicación de técnicas de aprendizaje automático, calculamos dos indicadores en tiempo real de la percepción de la inflación y mostramos que exhiben una dinámica comparable a la inflación y las expectativas de inflación, lo que sugiere que nuestros indicadores están estrechamente relacionados con los factores subyacentes que impulsan la percepción de la inflación entre enero de 2015 y marzo de 2023. En general, este enfoque proporciona un medio valioso para evaluar el sentimiento público hacia la inflación y ofrece una perspectiva complementaria a las medidas de expectativas de inflación tradicionales utilizadas en el marco de la política de inflación objetivo. |
| Keywords: | Inflation perceptions, Twitter, Real-time data, Central banks, Percepción de inflación, Twitter, medición en tiempo real, Bancos centrales. |
| JEL: | E31 E37 E52 |
| Date: | 2023–11 |
| URL: | http://d.repec.org/n?u=RePEc:bdr:borrec:1256&r=big |
| By: | Mariam Dundua (Financial and Supervisory Technology Development Department, National Bank of Georgia); Otar Gorgodze (Head of Financial and Supervisory Technologies Department, National Bank of Georgia) |
| Abstract: | The recent advances in Artificial Intelligence (AI), in particular, the development of reinforcement learning (RL) methods, are specifically suited for application to complex economic problems. We formulate a new approach looking for optimal monetary policy rules using RL. Analysis of AI generated monetary policy rules indicates that optimal policy rules exhibit significant nonlinearities. This could explain why simple monetary rules based on traditional linear modeling toolkits lack the robustness needed for practical application. The generated transition equations analysis allows us to estimate the neutral policy rate, which came out to be 6.5 percent. We discuss the potential combination of the method with state-of-the-art FinTech developments in digital finance like DeFi and CBDC and the feasibility of MonetaryTech approach to monetary policy. |
| Keywords: | Artificial Intelligence; Reinforcement Learning; Monetary policy |
| JEL: | C60 C61 C63 E17 C45 E52 |
| Date: | 2022–11 |
| URL: | http://d.repec.org/n?u=RePEc:aez:wpaper:2022-02&r=big |
| By: | Thomas R. Cook; Sophia Kazinnik; Anne Lundgaard Hansen; Peter McAdam |
| Abstract: | This study evaluates the performance of local large language models (LLMs) in interpreting financial texts, compared with closed-source, cloud-based models. We first introduce new benchmarking tasks for assessing LLM performance in analyzing financial and economic texts and explore the refinements needed to improve its performance. Our benchmarking results suggest local LLMs are a viable tool for general natural language processing analysis of these texts. We then leverage local LLMs to analyze the tone and substance of bank earnings calls in the post-pandemic era, including calls conducted during the banking stress of early 2023. We analyze remarks in bank earnings calls in terms of topics discussed, overall sentiment, temporal orientation, and vagueness. We find that after the banking stress in early 2023, banks tended to converge to a similar set of topics for discussion and to espouse a distinctly less positive sentiment. |
| Keywords: | data; large language models; quantitative methods; banking and finance |
| JEL: | C45 G21 |
| Date: | 2023–11–06 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedkrw:97255&r=big |
| By: | Grzegorz Marcjasz; Tomasz Serafin; Rafal Weron |
| Abstract: | We propose a novel electricity price forecasting model tailored to intraday markets with continuous trading. It is based on distributional deep neural networks with Johnson SU distributed outputs. To demonstrate its usefulness, we introduce a realistic trading strategy for the economic evaluation of ensemble forecasts. Our approach takes into account forecast errors in wind generation for four German TSOs and uses the intraday market to resolve imbalances remaining after day-ahead bidding. We argue that the economic evaluation is crucial and provide evidence that the better performing methods in terms of statistical error metrics do not necessarily lead to higher trading profits. |
| Keywords: | Intraday electricity market; Probabilistic forecast; Path forecast; Prediction bands; Trading strategy; Neural networks |
| JEL: | C22 C32 C45 C51 C53 Q41 Q47 |
| Date: | 2023 |
| URL: | http://d.repec.org/n?u=RePEc:ahh:wpaper:worms2301&r=big |
| By: | Seulki Chung |
| Abstract: | Feedforward neural network (FFN) and two specific types of recurrent neural network, long short-term memory (LSTM) and gated recurrent unit (GRU), are used for modeling US recessions in the period from 1967 to 2021. The estimated models are then employed to conduct real-time predictions of the Great Recession and the Covid-19 recession in US. Their predictive performances are compared to those of the traditional linear models, the logistic regression model both with and without the ridge penalty. The out-of-sample performance suggests the application of LSTM and GRU in the area of recession forecasting, especially for the long-term forecasting tasks. They outperform other types of models across 5 forecasting horizons with respect to different types of statistical performance metrics. Shapley additive explanations (SHAP) method is applied to the fitted GRUs across different forecasting horizons to gain insight into the feature importance. The evaluation of predictor importance differs between the GRU and ridge logistic regression models, as reflected in the variable order determined by SHAP values. When considering the top 5 predictors, key indicators such as the S\&P 500 index, real GDP, and private residential fixed investment consistently appear for short-term forecasts (up to 3 months). In contrast, for longer-term predictions (6 months or more), the term spread and producer price index become more prominent. These findings are supported by both local interpretable model-agnostic explanations (LIME) and marginal effects. |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2310.17571&r=big |
| By: | Hendrik Jenett |
| Abstract: | This study analyses the time-varying composition of real estate values by using an artificial neural network approach to identify whether and how certain indicators’ impacts on property values fluctuate over time. Therefore, cross-sectional property and macroeconomic data from the United States is applied, spanning a period from 1999 to 2021. In times of normal economic activity, property values are made up of two-thirds of physical attributes and one-third of the macroeconomic environment. During crises periods and times of high uncertainty, like the Global Financial Crisis, the share of the economies impact increases by roughly 5%, meaning that sudden economic changes have a higher impact on property values during crises periods versus normal times. However, these changes in the composition of real estate values varies even from one crisis to another, which confirms the dynamic relationship between the US macroeconomy and the housing market. Moreover, this study provides evidence that neural networks are capable of detecting non-linearities in property values especially during times of financial volatility. |
| Keywords: | Artificial Neural Network; Explainable Artificial Intelligence; Macroeconomy; Valuation |
| JEL: | R3 |
| Date: | 2023–01–01 |
| URL: | http://d.repec.org/n?u=RePEc:arz:wpaper:eres2023_183&r=big |
| By: | Chaohua Dong; Jiti Gao; Bin Peng; Yayi Yan |
| Abstract: | In this paper, we consider estimation and inference for both the multi-index parameters and the link function involved in a class of semiparametric multi-index models via deep neural networks (DNNs). We contribute to the design of DNN by i) providing more transparency for practical implementation, ii) defining different types of sparsity, iii) showing the differentiability, iv) pointing out the set of effective parameters, and v) offering a new variant of rectified linear activation function (ReLU), etc. Asymptotic properties for the joint estimates of both the index parameters and the link functions are established, and a feasible procedure for the purpose of inference is also proposed. We conduct extensive numerical studies to examine the finite-sample performance of the estimation methods, and we also evaluate the empirical relevance and applicability of the proposed models and estimation methods to real data. |
| Date: | 2023–11 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2311.02789&r=big |
| By: | Antonin Bergeaud; Cyril Verluise |
| Abstract: | We use patent data to study the contribution of the US, Europe, China and Japan to frontier technology using automated patent landscaping. We find that China's contribution to frontier technology has become quantitatively similar to the US in the late 2010s while overcoming the European and Japanese contributions respectively. Although China still exhibits the stigmas of a catching up economy, these stigmas are on the downside. The quality of frontier technology patents published at the Chinese Patent Office has leveled up to the quality of patents published at the European and Japanese patent offices. At the same time, frontier technology patenting at the Chinese Patent Office seems to have been increasingly supported by domestic patentees, suggesting the build up of domestic capabilities. |
| Keywords: | frontier technologies, China, patent landscaping, machine learning, patents |
| Date: | 2022–10–14 |
| URL: | http://d.repec.org/n?u=RePEc:cep:poidwp:039&r=big |
| By: | Bastian Krämer; Moritz Stang; Vanja Doskoc; Wolfgang Schäfers; Friedrich Tobias |
| Abstract: | The use of Automated Valuation Models (AVMs) in the context of traditional real estate valuations and their performance has been discussed in the academic community for several decades. Most studies focus on finding which method is best suited for estimating property values. One aspect that has not yet been studied scientifically is the appropriate choice of the spatial training level. The published research on AVMs usually deals with a manually defined region and fails to test the methods used on different spatial levels. The aim of our research is thus to investigate the impact of training AVM algorithms at different spatial levels in terms of valuation accuracy. We use a dataset with about 1.2 million residential properties from Germany and test four different methods, namely Ordinary Least Square, Generalized Additive Models, eXtreme Gradient Boosting and Deep Neural Network. Our results show that the right choice of spatial training level can have a major impact on the model performance, and that this impact varies across the different methods. |
| Keywords: | Automated Valuation Models; Machine Learning; Model Performance; Spatial Training Level |
| JEL: | R3 |
| Date: | 2023–01–01 |
| URL: | http://d.repec.org/n?u=RePEc:arz:wpaper:eres2023_120&r=big |
| By: | Kang Gao; Stephen Weston; Perukrishnen Vytelingum; Namid R. Stillman; Wayne Luk; Ce Guo |
| Abstract: | We propose the Chiarella-Heston model, a new agent-based model for improving the effectiveness of deep hedging strategies. This model includes momentum traders, fundamental traders, and volatility traders. The volatility traders participate in the market by innovatively following a Heston-style volatility signal. The proposed model generalises both the extended Chiarella model and the Heston stochastic volatility model, and is calibrated to reproduce as many empirical stylized facts as possible. According to the stylised facts distance metric, the proposed model is able to reproduce more realistic financial time series than three baseline models: the extended Chiarella model, the Heston model, and the Geometric Brownian Motion. The proposed model is further validated by the Generalized Subtracted L-divergence metric. With the proposed Chiarella-Heston model, we generate a training dataset to train a deep hedging agent for optimal hedging strategies under various transaction cost levels. The deep hedging agent employs the Deep Deterministic Policy Gradient algorithm and is trained to maximize profits and minimize risks. Our testing results reveal that the deep hedging agent, trained with data generated by our proposed model, outperforms the baseline in most transaction cost levels. Furthermore, the testing process, which is conducted using empirical data, demonstrates the effective performance of the trained deep hedging agent in a realistic trading environment. |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2310.18755&r=big |
| By: | Tomas Adam; Jan Belka; Martin Hluze; Jakub Mateju; Hana Prause; Jiri Schwarz |
| Abstract: | We use Mastercard card payments data to nowcast turnover in Czech retail sales and services. We show that an index based on this data tracks surprisingly well the official retail sales data released by the Czech Statistical Office (CZSO) more than a month later. We further show that the card payments data not only helps in backcasting Czech retail sales after the end of the month, but also provides valuable information for the nowcast as soon as three weeks into the ongoing month. That is six to seven weeks ahead of the official release. To illustrate the usefulness of our method, we show that we would have been able to backcast, with reasonable accuracy, the sharp drop in retail sales that occurred at the outbreak of the first wave of covid-19 in Czechia in March 2020 four weeks before the March data was released by the CZSO. |
| Keywords: | Card payments data, household consumption, household demand, nowcasting, retail sales, sales in services |
| JEL: | E21 E27 |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:cnb:wpaper:2023/14&r=big |
| By: | Tatiana Evdokimova (Joint Vienna Institute); Piroska Nagy Mohacsi (London School of Economics and Political Science); Olga Ponomarenko (Caplight); Elina Ribakova (Peterson Institute for International Economics) |
| Abstract: | This paper uses innovative natural language processing techniques to analyze central bank communication in emerging-market (EM) central banks and compare it with that of the Federal Reserve (Fed) and the European Central Bank (ECB). Once laggards of the central banking policy scene, EM central banks have made remarkable progress in improving their policy frameworks in the past two decades. They adopted many of the principles of advanced-economy (AE) central banks both in policy conduct and communication, but with modifications that reflect their specific circumstances of capital flow volatility, financial dollarization, and traditionally weaker credibility. The authors find that EM central banks' transparency has improved dramatically; their statements' readability has overall been better than in AEs; their focus on inflation has been sharper; and they have used data-shy "forward guidance" sparingly and flexibly. Worryingly though, most central banks do not communicate on inflationary pressures until after inflation already happens. EMs have outperformed AEs in two critical respects recently: addressing rising post-COVID inflationary pressures in a timely manner and, related, avoiding banking sector stress during the monetary policy tightening cycle. Systemic support in the form of currency swaps and repo operations by the Fed and the ECB with powerful signaling at times of acute market stress also helped. EM central banks have also started moving towards easing monetary policy already, ahead of the Fed and the ECB. Bringing down inflation fast and sustainably will be the ultimate test for the quality of EM central bank frameworks. The authors conclude with policy lessons for both EM and AE central banks. These include better forecasting and communication of inflation by the majority of central banks; more consistent delivery by EM central banks of communicated policy action; discarding pure "forward guidance" that hampers data dependency and thus fast policy action particularly at times of rapid change; consistent focus on supply-side factors of inflation; and for multiple-goal central banks, a clear choice and communication of policy priorities at times of possible conflict among some of the goals. The paper also suggests a more transparent communication of coordination with fiscal authorities that would improve the credibility of both the monetary and fiscal authorities. |
| Keywords: | central banking, monetary policy, emerging markets, Federal Reserve, ECB, communication, inflation-targeting, currency swaps, supply-side inflation, forward guidance, Chat GPT, AI |
| JEL: | B22 C55 E42 E52 E58 |
| Date: | 2023–10 |
| URL: | http://d.repec.org/n?u=RePEc:iie:wpaper:wp23-10&r=big |
| By: | Mamalis, Marios; Kalampokis, Evangelos; Karamanou, Areti; Brimos, Petros; Tarabanis, Konstantinos |
| Abstract: | Large language models possess tremendous natural language understanding and generation abilities. However, they often lack the ability to discern between fact and fiction, leading to factually incorrect responses. Open Government Data are repositories of, often times linked, information that is freely available to everyone. By combining these two technologies in a proof of concept designed application utilizing the GPT3.5 OpenAI model and the Scottish open statistics portal, we show that not only is it possible to augment the large language model's factuality of responses, but also propose a novel way to effectively access and retrieve statistical information from the data portal just through natural language querying. We anticipate that this paper will trigger a discussion regarding the transformation of Open Government Portals through large language models. |
| Date: | 2023–10–24 |
| URL: | http://d.repec.org/n?u=RePEc:osf:osfxxx:9b35z&r=big |