|
|
Text Classification Methods in the Context of Web Pages
Trstenský, Patrik ; Bartík, Vladimír (referee) ; Burget, Radek (advisor)
This work deals with the issue of text classification in the context of websites. It examines available classification methods and their accuracy over web page plain text. It deals with constructing a dataset for training these methods for a specific domain. We obtain data for creating the dataset from publicly available websites that utilize RDF documents defined in HTML code. The conclusion of the work consists of the creation of two datasets for two different domains. Furthermore, the use of these datasets for training models and testing of their accuracy.
|
|
|
Crude Oil Price Forecast based on Text News
Skalický, Jan ; Bojar, Ondřej (advisor) ; Žabokrtský, Zdeněk (referee)
For crude oil price forecast, there is a whole range of algorithms. In this thesis we bring out a new perspective on this issue and introduce our project COPF. Using a maximum entropy classifier, we try to predict the change in crude oil price from text information available on the Internet. We are taking advantage of the knowledge of experts in the field. As a part of the thesis, we tested and improved COPF precision. We have found out that this approach poses a lot of interesting problems. In the current state, the precision of our prediction surpassed the baseline but for further development, it is necessary to obtain more data sources. Our algorithm has never been regarded as a self-standing method but it may nicely complement numerical algorithms.
|
|
|
Popularity Meter
Hajič, Jan ; Bojar, Ondřej (advisor) ; Popel, Martin (referee)
Having the possibility of automatically tracking a person's popularity in the newspapers is an idea appealing not just to those in the media spotlight. While sentiment (subjectivity) analysis is a rapidly growing subfield of computational linguistics, no data from the news domain are yet available for Czech. We have therefore started building a manually annotated polarity corpus of sentences from Czech news texts; however, these texts have proven themselves rather unwieldy for such processing. We have also designed a classifier which should be able to track popularity based on this corpus; the classifier has been tested on a corpus of product reviews of domestic appliances and some introductory testing has been done on the nascent news corpus. As a model, we simply extract a unigram polarity lexicon from the data. We then use three related methods for identifying lemma polarity and a number of simple filters for feature selection. On the domestic appliance data, our simplest model has achieved results comparable to the state of the art, however, the properties of Czech news texts and preliminary results hint a more linguistically oriented approach might be preferrable.
|
|
|
Analýza textových používateľských hodnotení vybranej skupiny produktov
Valovič, Roman
This work focuses on the design of a system that identifies frequently discussed product features in product reviews, summarizes them, and displays them to the user in terms of sentiment. The work deals with the issue of natural language processing, with a specific focus on Czech languague. The reader will be introduced the methods of preprocessing the text and their impact on the quality of the analysis results. The identification of the mainly discussed products features is carried out by cluster analysis using the K-Means algorithm, where we assume that sufficiently internally homogeneous clusters will represent the individual features of the products. A new area that will be explored in this work is the representation of documents using the Word embeddings technique, and its potential of using vector space as input for machine learning algorithms.
|
|
|
Extraction of Semantic Relations from Text
Pospíšil, Milan ; Schmidt, Marek (referee) ; Smrž, Pavel (advisor)
Today exists many semi-structured documents, whitch we want convert to structured form. Goal of this work is create a system, that make this task more automatized. That could be difficult problem, because most of these documents are not generated by computer, so system have to tolerate differences. We also need some semantic understanding, thats why we choose only domain of meeting minutes documents.
|
|
|
Comparison of approaches to text classification
Knížek, Jan ; Hana, Jiří (advisor) ; Vidová Hladká, Barbora (referee)
The focus of this thesis is short text classification. Short text is the prevailing form of text on e-commerce and review platforms, such as Yelp, Tripadvisor or Heureka. As the popularity of the online communication is increasing, it is becoming infeasible for users to filter information manually. It is therefore becoming more and more important to recog- nise the relevant information in text. Classification of reviews is especially challenging, because they have limited structure, use informal language, contain a high number of errors and rely heavily on context and common knowledge. One of the possible appli- cations of machine learning is to automatically filter data and show users only relevant pieces of information. We work with restaurant reviews from Yelp and aim to predict their usefulness. Most restaurants have relatively many reviews, yet only few are truly useful. Our objective is to compare machine learning methods for predicting usefulness. 1
|
|
|
Artificial Intelligence Document Classification
Molnár, Ondřej ; Kačic, Matej (referee) ; Třeštíková, Lenka (advisor)
This paper deals with document classification using artificial intelligence. It describes the principles of classification and machine learning. It also introduces AI methods and presents Naive Bayes classification method in detail. Provides practical implementation of the classifier in MS Office and discusses other possible extensions.
|
|
|
Scala Programming Language and Its Use for Data Analysis
Kohout, Tomáš ; Bartík, Vladimír (referee) ; Zendulka, Jaroslav (advisor)
This thesis deals with comparing the Scala programming language with other commonly used languages for data analysis. These languages are evaluated on the basis of the following categories: data manipulation and visualization, machine learning and concurent processing capabilities. The evaluation then shows the strengths and weaknesses of Scala. The strengths will be demonstrated on application for email categorization.
|
|
|
Recurrent Neural Network for Text Classification
Myška, Vojtěch ; Kolařík, Martin (referee) ; Povoda, Lukáš (advisor)
Thesis deals with the proposal of the neural networks for classification of positive and negative texts. Development took place in the Python programming language. Design of deep neural network models was performed using the Keras high-level API and the TensorFlow numerical computation library. The computations were performed using GPU with support of the CUDA architecture. The final outcome of the thesis is linguistically independent neural network model for classifying texts at character level reaching up to 93,64% accuracy. Training and testing data were provided by multilingual and Yelp databases. The simulations were performed on 1200000 English, 12000 Czech, German and Spanish texts.
|
|
|
Popularity Meter
Hajič, Jan ; Bojar, Ondřej (advisor) ; Popel, Martin (referee)
Having the possibility of automatically tracking a person's popularity in the newspapers is an idea appealing not just to those in the media spotlight. While sentiment (subjectivity) analysis is a rapidly growing subfield of computational linguistics, no data from the news domain are yet available for Czech. We have therefore started building a manually annotated polarity corpus of sentences from Czech news texts; however, these texts have proven themselves rather unwieldy for such processing. We have also designed a classifier which should be able to track popularity based on this corpus; the classifier has been tested on a corpus of product reviews of domestic appliances and some introductory testing has been done on the nascent news corpus. As a model, we simply extract a unigram polarity lexicon from the data. We then use three related methods for identifying lemma polarity and a number of simple filters for feature selection. On the domestic appliance data, our simplest model has achieved results comparable to the state of the art, however, the properties of Czech news texts and preliminary results hint a more linguistically oriented approach might be preferrable.
|
guest ::
