|
| |
|
|
News Feed Classifications to Improve Volatility Predictions
Pogodina, Ksenia ; Šopov, Boril (advisor) ; Červinka, Michal (referee)
This thesis analyzes various text classification techniques in order to assess whether the knowledge of published news articles about selected companies can improve its' stock return volatility modelling and forecasting. We examine the content of the textual news releases and derive the news sentiment (po larity and strength) employing three different approaches: supervised machine learning Naive Bayes algorithm, lexicon-based as a representative of linguistic approach and hybrid Naive Bayes. In hybrid Naive Bayes we consider only the words contained in the specific lexicon rather than whole set of words from the article. For the lexicon-based approach we used independently two lexicons one with binary another with multiclass labels. The training set for the Naive Bayes was labeled by the author. When comparing the classifiers from the machine learning approach we can conclude that all of them performed similarly with a slight advantage of the hybrid Naive Bayes combined with multiclass lexicon. The resulting quantitative data in form of sentiment scores will be then incorpo rated into GARCH volatility modelling. The findings suggest that information contained in news feeds does bring an additional explanatory power to tradi tional GARCH model and is able to improve it's forecast. On the...
|
|
|
Multi-source Text Summarization for Czech
Brus, Tomáš ; Bojar, Ondřej (advisor) ; Mareček, David (referee)
This work focuses on the summarization task for a set of articles on the same topic. It discusses several possible ways of summarizations and ways to assess their final quality. The implementation of the described algorithms and their application to selected texts constitutes a part of this work. The input texts come from several Czech news servers and they are represented as deep syntactic trees (the so called tectogrammatical layer).
|
|
|
Předpovídání trendů akciového trhu z novinových článků
Serebryannikova, Anastasia ; Kuboň, Vladislav (advisor) ; Vidová Hladká, Barbora (referee)
In this work we made an attempt to predict the upwards/downwards movement of the S&P 500 index from the news articles published by Bloomberg and Reuters. We employed the SVM classifier and conducted multiple experiments aiming at understanding the shape of the data and the specifics of the task better. As a result, we established the common evaluation settings for all our subsequent experiments. After that we tried incorporating various features into the model and also replicated several approaches previously suggested in the literature. We were able to identify some non-trivial dependencies in the data which helped us achieve a high accuracy on the development set. However, none of the models that we built showed comparable performance on the test set. We have come to the conclusion that whereas some trends or patterns can be identified in a particular dataset, such findings are usually barely transferable to other data. The experiments that we conducted support the idea that the stock market is changing at random and a high quality of prediction may only be achieved on particular sets of data and under very special settings, but not for the task of stock market prediction in general. 1
|
|
|
The Case Cervanova: from reality, newspaper news to feature film and film reviews.
Zajíčková, Monika ; Novotný, David Jan (advisor) ; Štoll, Martin (referee)
The master thesis "The Case Cervanova: from reality and newspaper to feature film and film reviews" deals with a criminal case that happened in the summer of 1976. Young student Ľudmila Cervanová was kidnapped, raped and killed. The perpetrators were caught in 1981 and the judicial process lasted until 2006. Several literary works have been written and two films have been made, Bolero and Kauza Cervanová. The work focuses on literary works and on both films where it examines the context in which the authors presented the story and how they worked with the facts. In both films, emphasis is also placed on how they were received and judged by the media and the public, thus the work examines the reviews and opinions of the audience. Emphasis is also placed on newspaper of that time and their interpretation of the case. The research sample contains all the relevant dates. That is, from June 1981 to October 1982 (the period of detention and conviction of the offenders), together with April 1983 (decision of the appeal court). The research sample of this period includes four main daily newspapers of that time: Mladá fronta: deník mladých lidí; Lidová demokracie: Orgán československé strany lidové; Rudé právo: Orgán Československé sociálně demokratické strany dělnické a Svobodné slovo. Another relevant...
|
|
|
News Feed Classifications to Improve Volatility Predictions
Pogodina, Ksenia ; Šopov, Boril (advisor) ; Červinka, Michal (referee)
This thesis analyzes various text classification techniques in order to assess whether the knowledge of published news articles about selected companies can improve its' stock return volatility modelling and forecasting. We examine the content of the textual news releases and derive the news sentiment (po larity and strength) employing three different approaches: supervised machine learning Naive Bayes algorithm, lexicon-based as a representative of linguistic approach and hybrid Naive Bayes. In hybrid Naive Bayes we consider only the words contained in the specific lexicon rather than whole set of words from the article. For the lexicon-based approach we used independently two lexicons one with binary another with multiclass labels. The training set for the Naive Bayes was labeled by the author. When comparing the classifiers from the machine learning approach we can conclude that all of them performed similarly with a slight advantage of the hybrid Naive Bayes combined with multiclass lexicon. The resulting quantitative data in form of sentiment scores will be then incorpo rated into GARCH volatility modelling. The findings suggest that information contained in news feeds does bring an additional explanatory power to tradi tional GARCH model and is able to improve it's forecast. On the...
|
|
|
Emigration from the Canary Islands to Cuba, 1860-1914
Špitálská, Lucie ; Křížová, Markéta (advisor) ; Opatrný, Josef (referee)
(in English): The aim of this thesis is to introduce the theme of the Canary emigration to Cuba in the second half of the nineteenth and early twentieth century. For better orientation, the author has chosen time interface between the years 1860-1914. The introduction outlines the sources and secondary literature, from which the author draws and also supposed goal of work. The second chapter is devoted to the Canarian emigration from Columbus's expeditions to the first half of the nineteenth century and its gradual development. The third chapter describes the specific reasons that led to the emigration from the Canary Islands, the attention is focused on the travel conditions of immigrants, employment contracts and promotion. The fourth chapter has been devoted specifically Canarian immigrants and their impact on Cuba, including labor stratification. The fifth chapter focuses on emigration to the turn of the century, there is a special attention is paid to the topic canary societies and associations. The conclusion summarizes the achieved knowledges and changes of this emigration.
|
|
|
Multi-source Text Summarization for Czech
Brus, Tomáš ; Bojar, Ondřej (advisor) ; Mareček, David (referee)
This work focuses on the summarization task for a set of articles on the same topic. It discusses several possible ways of summarizations and ways to assess their final quality. The implementation of the described algorithms and their application to selected texts constitutes a part of this work. The input texts come from several Czech news servers and they are represented as deep syntactic trees (the so called tectogrammatical layer).
|
|
|
Death in the media nowadays and twenty years ago
STAVRAKI, Arina
This bachelor thesis deals with death in media nowadays and twenty years ago. Some articles were chosen from Lidové noviny and Mladá fornta DNES newspapers issued in 1990, 1991, 2010 and 2011. Those articles described tragic events when somebody died. These topics were divided into the following subcategories: death in a car accident, train collisions, earthquakes, murders, suicides, death of an important person, restoring the death penalty, abortion and euthanasia. The chosen articles were afterwards analysed and compared. Firstly, the thesis analyses 1990 and 1991 articles and consequently evaluates this chosen period. Then followed the analysis of the 2010 and 2011 articles. The final evaluation of the work consists of the particular comparison of the articles on the same, similar or completely different events within twenty years. This bachelor thesis maps view changes of these events in journalism within the given period in terms of style of informing readers about tragic events and also the increased rate of death reporting in general
|
|
|
The picture of life of disabled individuals in selected Czech media
BÍLKOVÁ, Ivana
The work deals with the life's portrait of disabled persons in selected Czech media. The theoretical part provides clarification of basic terminology data, such as disability, physiotherapy and integration. Particular attention is paid to integration of the individuals into the society. Furthermore, the development of the picture of disabled persons' life trough the history is integrated here ? in the ancient world, Middle Ages and modern period. A chapter dealing with the characteristics of media, their functions and effects is also included to this work. The practical part is focused on analyses of the occurrence and the content of articles about disabled individuals from two picked newspapers ? MF Dnes as a representative of the national Czech daily press and Pelhřimovký deník as a representative of the regional press. The research is carried out during the monitored period from July to December 2010. Positive attitude to handicapped individuals and groups as well as openness to their integration into the general population is obvious from the analyzed newspaper articles.
|
guest ::
