Relationship between e-reputation and stock performance: evidence from Turkish airline industry
Yükleniyor...
Dosyalar
Tarih
2021
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
İstanbul Ticaret Üniversitesi
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
Abstract Determinants of stock market behavior have always interested researchers and has been studied widely, however development of information and communication technologies and availability of daily fluctuating public data created new areas for research. This study analyzes impact of 2 variables on stock price of 2 airline companies which are traded at Istanbul Stock Exchange (BIST). Variables are followings; (1) public tweets mentioning company twitter account or traded stock name, (2) Google search volume about the company or its stock name. Using manual labeling, decision trees and deep learning techniques semantic analysis has been performed on 34,152 tweets. Aside that 78. 866 tweets mentioning companies’ stock names has been analyzed for the duration of 6 months using conventional methods. Google search volume about mentioned 2 companies and their stock name was studied for the period of 2. 5 years. It was found that daily share tweets likes has moderate positive correlation with daily stock change and daily traded stock volume. Semantic tweets displayed no statistically significant correlation with stock performance. For Google Trends, search for company share has moderate positive correlation with daily stock volume for one airline company and search for company name had low positive correlation with daily stock close price for another airline company. And lastly correlation between two company’s Twitter Activities, Google Search Volumes, and Stock performances were analyzed to see if their was any apparent industry trend and whether company’s reputation and stock performance move together. For all three platforms the answer was yes. Keywords: E-reputation,Semantic Analysis, Twitter, Google Trends, Stock, Borsa Istanbul Table of Contents Table list………………………………………………………………………………. 5 Picture list………………………………………………………………………………. 5 Abbreviations………………………………………………………………………………. 6 1 Introduction………………………………………………………………………………. 7 2 Literature Review…………………………………………………………………………. 9 3 Data Collection and Methodology…………………………………………………………. 12 3. 1 Twitter Data Collection……………………………………………………………. …. 12 3. 2 Data Preprocessing for Share Tweets……………………………………………. ……15 3. 3 Manual Labeling of Semantic Tweets…………………………………………………. 17 3. 4 Preprocessing data before applying deep learning models……………………………. 21 3. 5 Applying Deep Learning Models………………………………………………………. 23 3. 6 Applying Decision Tree…………………………………………………………………28 Data Collection for Google Search Volume…………………………………………………35 4 Results…………………………………………………………………………………………35 5 Discussion……………………………………………………………………………………. 56 5. 1 Discussion about share tweets and stock performance…………………………………. 56 5. 2 Discussion about semantic tweets and stock performance………………………………. 58 5. 3 Discussion about Google Search Volume and Stock Performance……………………. 61 5. 4 Discussion About Stock Performance of Two Companies………………………………. 63 5. 5 Qualitative Findings from Manual Labeling of Eight Thousand Tweets………………. 64 5. 5. 1 Services that Affect Consumers Most………………………………………………. 64 5. 5. 1. 1 Pricing Strategy of Company and Fin-tech solutions…………………………. 64 5. 5. 1. 2 Customer Support Channels……………………………………………………. 65 5. 5. 1. 3 Stuff Attitude……………………………………………………………………67 5. 5. 1. 4 Ensuring security of customers property………………………………………. 67 5. 5. 1. 5 Flight delay or cancellation……………………………………………………. 67 5. 5. 1. 6 Information Related Problems…………………………………………………. 68 5. 5. 1. 7 Service Distributors and Partners………………………………………………. 68 5. 5. 1. 8 Reacting to Covid-19……………………………………………………………68 5. 5. 1. 9 Flight Experience………………………………………………………………. 69 5. 5. 2 Airline Service Provision in 21st century……………………………………………69 5. 5. 2. 1 Managing E-Reputation…………………………………………………………………69 5. 5. 2. 2 NLP Challenges in Business Intelligence Solutions……………………………. 70 5. 5. 2. 3 Software Integration With Third-Party Service Providers………………………70 5. 5. 2. 4 Challenges of using Twitter as Ticketing Support System………………………71 5. 5. 2. 5 Robot Investors and Financial Analysts recommendations……………………. 71 6 Conclusion……………………………………………………………………………………. 71 7 References………………………………………………………………………………………73
Açıklama
Tez (Yüksek Lisans) -- İstanbul Ticaret Üniversitesi -- Kaynakça var.
