|
|
Sentiment Analysis for the Field of Computer Games
Balajka, Pavel ; Otrusina, Lubomír (referee) ; Smrž, Pavel (advisor)
The thesis deals with sentiment analysis extracted from opinions of users on social \mbox{networks}. It describes a general system that was created for presented purpose and specialised on the field of strategic computer games. In particular we unravel the problems of acquiring data from social networks, sentiment analysis and results presentation to the user. We mention particular ways of text processing e.g. tokenization and unnecessary word filtration, for purpose of more effective sentiment analysis and we mention machine learning methods e.g. Decision Trees and Naive Bayes, and their usage. Next we describe design of desired system and its implementation with chosen parts and methods. In the end we compare results of tests of sentiment analyzator done under various circumstances.
|
|
|
Cooperation of Multiagent Group in Dynamic Environment
Balajka, Pavel ; Uhlíř, Václav (referee) ; Zbořil, František (advisor)
The thesis summarises thorough analysis of rules and mechanisms of Multiagent Contest for year 2020 followed by design of multiagent group which will try to succeed in a simulation of the mentioned contest with the best results. The design was created according to the Prometheus metodology from general view (scenarios, goals) to specific details (agent plans, messages). Implementation is realized through JaCaMo system, which is also briefly described in this paper. Then the pros and cons of JaCaMo system are analyzed. Lastly the implemented solution of this work is compared with other teams that participated in the mentioned agent competition this year.
|
|
|
Web App for Management of Appointments
Balajka, Pavel ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
The aim of this work is to create an aplication that allows lectors create and manage consultation appointments on which students will be able to sing on. Therefore a web application designed to run on a server was created, that lectors and students can access with an internet browser. The mentioned problem is solved in a way which simplifies the work of its users and offers them additional possibilities regarding consultation management. For example, the application is able to send a notification when a section of a consultation becomes available or to show consultation's history - which user has done what action with a chosen consultation. The procedure of simple usage of the application is the following: Lector creates a consultation - chooses date, time and number of sections. Student selects a lector in the appliacion, displays his consultations, chooses desired time and signs on the consultation by just one click. Application was developed primarily for university lectors and students but it can be as well used by other subjects that need or want a system where they can create an event on which other people can sign on.
|
|
|
Cooperation of Multiagent Group in Dynamic Environment
Balajka, Pavel ; Uhlíř, Václav (referee) ; Zbořil, František (advisor)
The thesis summarises thorough analysis of rules and mechanisms of Multiagent Contest for year 2020 followed by design of multiagent group which will try to succeed in a simulation of the mentioned contest with the best results. The design was created according to the Prometheus metodology from general view (scenarios, goals) to specific details (agent plans, messages). Implementation is realized through JaCaMo system, which is also briefly described in this paper. Then the pros and cons of JaCaMo system are analyzed. Lastly the implemented solution of this work is compared with other teams that participated in the mentioned agent competition this year.
|
|
|
Sentiment Analysis for the Field of Computer Games
Balajka, Pavel ; Otrusina, Lubomír (referee) ; Smrž, Pavel (advisor)
The thesis deals with sentiment analysis extracted from opinions of users on social \mbox{networks}. It describes a general system that was created for presented purpose and specialised on the field of strategic computer games. In particular we unravel the problems of acquiring data from social networks, sentiment analysis and results presentation to the user. We mention particular ways of text processing e.g. tokenization and unnecessary word filtration, for purpose of more effective sentiment analysis and we mention machine learning methods e.g. Decision Trees and Naive Bayes, and their usage. Next we describe design of desired system and its implementation with chosen parts and methods. In the end we compare results of tests of sentiment analyzator done under various circumstances.
|
|
|
Web App for Management of Appointments
Balajka, Pavel ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
The aim of this work is to create an aplication that allows lectors create and manage consultation appointments on which students will be able to sing on. Therefore a web application designed to run on a server was created, that lectors and students can access with an internet browser. The mentioned problem is solved in a way which simplifies the work of its users and offers them additional possibilities regarding consultation management. For example, the application is able to send a notification when a section of a consultation becomes available or to show consultation's history - which user has done what action with a chosen consultation. The procedure of simple usage of the application is the following: Lector creates a consultation - chooses date, time and number of sections. Student selects a lector in the appliacion, displays his consultations, chooses desired time and signs on the consultation by just one click. Application was developed primarily for university lectors and students but it can be as well used by other subjects that need or want a system where they can create an event on which other people can sign on.
|
guest ::
