Dumitru Alexandru Mara

Mr. Dumitru Alexandru Mara

Lucian Blaga University of Sibiu, Sibiu, Romania
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 November 2022 in Mathematics
Verified
Journal Article
Open access
PDF file
Marian Pompiliu I CristescuMarian Pompiliu I Cristescu
Marian Pompiliu I Cristescu
Raluca Andreea NerișanuRaluca Andreea Nerișanu
Raluca Andreea Nerișanu
Dumitru Alexandru MaraDumitru Alexandru Mara
Dumitru Alexandru Mara
Simona OpreaSimona Oprea
Simona Oprea
(1) Background: Since the current crises that has inevitably impacted the financial market, market prediction has become more crucial than ever. The question of how risk managers can more accurately predict the evolution of their portfolio, while taking into consideration systemic risks brought on by a systemic crisis, is raised by the low rate of success of portfolio risk-management models. Sentiment analysis on natural language sentences can increase the accuracy of market prediction because financial markets are influenced by investor sentiments. Many investors also base their decisions on information taken from newspapers or on their instincts. (2) Methods: In this paper, we aim to highlight how sentiment analysis can improve the accuracy of regression models when predicting the evolution of the opening prices of some selected stocks. We aim to accomplish this by comparing the results and accuracy of two cases of market prediction using regression models with and without market news sentiment analysis. (3) Results: It is shown that the nonlinear autoregression model improves its goodness of fit when sentiment analysis is used as an exogenous factor. Furthermore, the results show that the polynomial autoregressions fit better than the linear ones. (4) Conclusions: Using the sentiment score for market modelling, significant improvements in the performance of linear autoregressions are showcased.

19 October 2023 in Information
Verified
Journal Article
Open access
PDF file
Marian Pompiliu I CristescuMarian Pompiliu I Cristescu
Marian Pompiliu I Cristescu
Dumitru Alexandru MaraDumitru Alexandru Mara
Dumitru Alexandru Mara
CLCulda Lia
Culda Lia
Raluca Andreea NerișanuRaluca Andreea Nerișanu
Raluca Andreea Nerișanu
Adela BaraAdela Bara
Adela Bara
Simona OpreaSimona Oprea
Simona Oprea
This study explores the potential of data science software solutions like Customer Relationship Management Software (CRM) for increasing the revenue generation of businesses. We focused on those businesses in the accommodation and food service sector across the European Union (EU). The investigation is contextualized within the rising trend of data-driven decision-making, examining the potential correlation between data science applications and business revenues. By employing a comprehensive evaluation of Eurostat datasets from 2014 to 2021, we used both univariate and multivariate analyses, assessing the percentage of companies that have e-commerce sales across the EU countries, focusing on the usage of big data analytics from any source and the use of CRM tools for marketing purposes or other activities. Big data utilization showed a clear, positive relationship with enhanced e-commerce sales. However, CRM tools exhibited a dualistic impact: while their use in marketing showed no significant effect on sales, their application in non-marketing functions had negative effects on sales. These findings underscore the potential role of CRM and data science solutions in enhancing business performance in the EU’s accommodation and food service industry.

30 November 2023 in Mathematics
Verified
Journal Article
Open access
PDF file
Marian Pompiliu I CristescuMarian Pompiliu I Cristescu
Marian Pompiliu I Cristescu
Dumitru Alexandru MaraDumitru Alexandru Mara
Dumitru Alexandru Mara
Raluca Andreea NerișanuRaluca Andreea Nerișanu
Raluca Andreea Nerișanu
CLCulda Lia
Culda Lia
IMIonela Maniu
Ionela Maniu
This study investigates the complex interplay between public sentiment, as captured through news titles and descriptions, and the stock prices of three major tech companies: Microsoft (MSFT), Tesla (TSLA), and Apple (AAPL). Leveraging advanced analytical methods including Pearson correlation, wavelet coherence, and regression analysis, this research probes the degree to which stock-price fluctuations can be attributed to the polarity of media sentiment. The methodology combines statistical techniques to assess sentiment’s predictive power for stock opening and closing prices, while wavelet coherence analysis unveils the temporal dynamics of these relationships. The results demonstrate a significant correlation between sentiment polarity and stock prices, with description polarity affecting Microsoft’s opening prices, title polarity influencing Tesla’s opening prices, and a positive impact of title polarity on Apple’s closing prices. However, Tesla’s stock showed no significant coherence, indicating a potential divergence in how sentiment affects stock behavior across companies. The study highlights the importance of sentiment analysis in forecasting stock-market trends, revealing not only direct correlations but also lagged influences on stock prices. Despite its focus on large-cap tech firms, this research provides a foundational understanding of sentiment’s financial implications, suggesting further investigation into smaller firms and other market sectors.

1 December 2022 in Journal of Social and Economic Statistics
Verified
Journal Article
Open access
PDF file
MCMarian Pompiliu Cristescu
Marian Pompiliu Cristescu
RNRaluca Andreea Nerişanu
Raluca Andreea Nerişanu
Dumitru Alexandru MaraDumitru Alexandru Mara
Dumitru Alexandru Mara
Sentiment analysis refers to the analysis of human opinions and sentiments that are expressed in written text, being also a part of the Natural Language Processing (NLP) tasks. Sentiment analysis can be applied in different domains, especially in the corporate marketing and sales, the healthcare system or the financial market analysis. In this paper we aim to highlight how data mining is able to extract the sentiment score from a financial platform that shows the major headlines regarding stocks, in order to highlight the publications’ positive or negative opinion over a stock. In order to gain the sentiment score we have scraped text data from the platform Finviz from which the polarity of the opinion may be extracted. We have also used Valence Aware Dictionary for Sentiment Reasoning (VADER), by running a Python script using the BeautifulSoup library. After that we have used Pandas (Python Data Analysis Library) to analyse and obtain a sentiment score on the article headlines. Results show that the script is able to generate the sentiment score for various selected stocks, while also showing graphical diagrams for the past and future trend of the stock, in terms of overall opinion on the stock performance.

1 November 2023 in Preprints.org
Verified
Preprint
Open access
PDF file
Marian Pompiliu I CristescuMarian Pompiliu I Cristescu
Marian Pompiliu I Cristescu
Dumitru Alexandru MaraDumitru Alexandru Mara
Dumitru Alexandru Mara
Raluca Andreea NerișanuRaluca Andreea Nerișanu
Raluca Andreea Nerișanu
CLCulda Lia
Culda Lia
IMIonela Maniu
Ionela Maniu
This study investigates the complex interplay between public sentiment, as captured through news titles and descriptions, and the stock prices of three major tech companies: Microsoft (MSFT), Tesla (TSLA), and Apple (AAPL). Leveraging advanced analytical methods including Pearson correlation, wavelet coherence, and regression analysis, this research probes the degree to which stock price fluctuations can be attributed to the polarity of media sentiment. The methodology combines statistical techniques to assess sentiment's predictive power on stock opening and closing prices, while wavelet coherence analysis unveils the temporal dynamics of these relationships. Results demonstrate a significant correlation between sentiment polarity and stock prices, with description polarity affecting Microsoft's opening prices, title polarity influencing Tesla's opening prices, and a positive impact of title polarity on Apple's closing prices. However, Tesla's stock showed no significant coherence, indicating a potential divergence in how sentiment affects stock behavior across companies. The study highlights the importance of sentiment analysis in forecasting stock market trends, revealing not only direct correlations but also lagged influences on stock prices. Despite its focus on large-cap tech firms, this research provides a foundational understanding of sentiment's financial implications, suggesting further investigation into smaller firms and other market sectors.

27 November 2025 in Electronics
Verified
Journal Article
Open access
PDF file
Marian Pompiliu I CristescuMarian Pompiliu I Cristescu
Marian Pompiliu I Cristescu
CBClaudiu Brandas
Claudiu Brandas
Dumitru Alexandru MaraDumitru Alexandru Mara
Dumitru Alexandru Mara
IPIoana Petrea
Ioana Petrea
The increasing use of complex “black-box” models for financial news sentiment analysis presents a challenge in high-stakes settings where transparency and trust are paramount. This study introduces and validates a finance-focused, fully reproducible, open-source workflow for building, explaining, and evaluating sector-specific sentiment models mapped to standard market taxonomies and investable proxies. We benchmark interpretable and transformer-based models on public datasets and a newly constructed, manually annotated gold-standard corpus of 1500 U.S. sector-tagged financial headlines. While a zero-shot FinBERT establishes a reasonable baseline (macro F1 = 0.555), fine-tuning on our gold data yields a robust macro F1 = 0.707, a substantial uplift. We extend explainability to the fine-tuned FinBERT with Integrated Gradients (IG) and LIME and perform a quantitative faithfulness audit via deletion curves and AOPC; LIME is most faithful (AOPC = 0.365). We also quantify the risks of weak supervision: accuracy drops (−21.0%) and explanations diverge (SHAP rank ρ = 0.11) relative to gold-label training. Crucially, econometric tests show the sentiment signal is reactive, not predictive, of next-day returns; yet it still supports profitable sector strategies (e.g., Technology long-short Sharpe 1.88). Novelty lies in a finance-aligned, sector-aware, trustworthiness blueprint that pairs fine-tuned FinBERT with audited explanations and uncertainty checks, all end-to-end reproducible and tied to investable sector ETFs.