|
|
User preference visualization for music
Gajdušek, Pavel ; Peška, Ladislav (advisor) ; Škoda, Petr (referee)
Most of the music portals offer users lists of songs that are the result of black-box algorithms. The recommendation is often nontransparent for users, therefore the irrele- vant recommendation might have negative consequences. The recommendation is mainly based on the computation of similarities between users or objects. The computation relies on collaborative techniques or similarity of the contents of the objects. The purpose of this bachelor thesis is to design and implement suitable visualization of these relations in the form of an interactive graph for a certain Spotify user. The visualization should help users realize that their data have inner structures and the recommendations are based on them. The final application should also provide a music playback using the songs contained in the graph. 1
|
|
|
Framework Supporting Online Evaluation of Recommender Systems
Novák, Ondřej ; Peška, Ladislav (advisor) ; Balcar, Štěpán (referee)
This work aims to highlight the importance of online evaluation for the testing of recommender systems. Firstly, we will look at the methods and the ways modern recommender systems operate. We will also introduce how they are compared both in online and offline settings. With this knowledge, we aim to build a .NET framework capable of tracking the various recommender systems for the purpose of measuring and comparing their performance during online use. To showcase the functionality of this framework, we use it to create a mock-up of an online movie database, where users can rate movies and receive movie recommendations.
|
|
|
Recommender System for Web Articles
Kočí, Jan ; Kesiraju, Santosh (referee) ; Fajčík, Martin (advisor)
Tématem této bakalářské práce jsou doporučovací systémy pro webové články. Tato práce nejdříve uvádí nejpopulárnější metody z této oblasti a vysvětluje jejich principy, následně navrhuje požití vlastní architektury, založené na neuronových sítích, která aplikuje metodu Skip-gram negative sampling na problematiku doporučování. V další části pak implementuje tuto architekturu společně s několika dalšími modely, požívající algoritmus SVD, collaborative filtering s algoritmem ALS a také metodu Doc2Vec k vytvoření vektorové reprezentace z obsahu získaných článků. Na závěr vytváří tři evaluační metriky, konkrétně metriky RANK, Recall at k a Precision at k, a vyhodnocuje kvalitu implementovaných modelů srovnáním výsledků s nejmodernějšími modely. Kromě toho také diskutuje o roli a smyslu doporučovacích systémů ve společnosti a uvádí motivaci pro jejich používání.
|
|
|
Educational System for Recommending Study Activities
Zapletal, Jakub ; Bartík, Vladimír (referee) ; Burget, Radek (advisor)
Cílem této práce je navrhnout a implementovat modul do existujícího doporučovacího systému Open University v Milton Keynes. Nyní nasazený doporučovací systém využívá informací o aktivitě uživatelů ve Virtual Learning Environment (VLE) nasbíraných z předchozích let a podle ní doporučuje studentům relevantní studijní aktivity. Tento modul využívá sémantické podobnosti mezi studijními materiály k doporučení těch, které pomohou uživateli vyřešit úkol nebo které jsou podobné k těm, o něž projevil zájem.K počítání podobnosti dokumentů je využíváno metod Term Frequency - Inverse Document Frequency a vnoření slov.Pro používání modulu a jeho komunikaci modulu s rozhraním OU Analyse je implementováno RESTful API.
|
|
|
Music recommendation based on music information retrieval
Semela, René ; Schimmel, Jiří (referee) ; Kiska, Tomáš (advisor)
This thesis deals with the design, implementation and testing of the content-based music recommender system based on music information retrieval. In the introduction the attention is paid to issues of music information retrieval and to areas of their utilization, it also focuses on tools of their retrieving. Aferwards the most used types of recommender systems are described, including their typical problems. Options of the hybridization of these systems as well as examples of the popular music recommender systems are mentioned. There also is an outline of their functioning. The following section is focused on the parameterization of musical pieces and is devoted to the description of particular most used parameters. The next section is devoted to the content-based music recommender system design itself, including the defining of particular parameters that are used to differentiate musical recordings using the algorithm mRMR and other procedures. The recommender system design as such is oriented to the classification method k-nearest neighbors. The attention is also paid to the model of user taste recorded by Rocchio algorithm. In the next section the system is implemented according to the design. There is also described its functionality including the background processes. The final part of this work is focused on system testing and evaluation.
|
|
|
Recommender systems for culture events
Vytisková, Zuzana ; Peška, Ladislav (advisor) ; Kopecký, Michal (referee)
The diploma thesis deals with the topic of recommendation in culture. In the theoretical part, it compares the recommendation of digitally available works with event recommendations, which serves as the basis for describing recommendations on the cultural portal. Further, the thesis examines the domain model as several different interconnected types of objects. Using these relations to enrich data sets allows overcoming the low data density and improving the recommendations. The paper examines two common situations of practical recommendation, general user recommendation with minimal profile and recommendation to registered users with known history. As a part of the solution, hybrid algorithms have been implemented based on the introducing content information into existing collaborative filtering methods. The results are verified in offline tests on data sets consisting of both research and real-world data. The subjective quality of the resulting recommendations was examined through a user study.
|
|
|
Methodology of implementing a recommender system using the Soyka tool
Müller, Petr ; Gála, Libor (advisor) ; Fanta, Michal (referee)
This diploma thesis is focused on content personalization and specifically on recommender systems. The aim of the thesis is to propose a methodology of recommender system implementation in e-commerce using the IT tool Soyka. Functions of personalization tools which directly support recommender systems are identified on the basis of a theoretical description of recommender systems and their technological approaches. Based on these identified functions the tool Soyka is classified. The main contribution of the thesis is the created and published methodology which is ready to be used on real implementation projects by anyone involved.
|
|
|
Books Recommender System via Linked Open Data
Maleček, Ladislav ; Peška, Ladislav (advisor) ; Škoda, Petr (referee)
This thesis focuses on using recommender system's methods on Linked Open Data in a domain of books. After thorough analysis of multiple available Linked Open Data sets, we have concluded that data sets of sufficient size and quality already exist. Together with careful analysis of the structure and quality of the data, recommender system web application has been developed based on retrieved data from a Wikidata endpoint. The application design allows an incorporation of data from multiple sources. A novel approach for generating recommendations utilizing multi language tags extracted from Wikipedia was used. We have shown that it is possible and viable to use recommender systems on top of the Linked Open Data, but the common recommender system's algorithms have to be modified in order to deal with a huge amount of sparsity in the data.
|
|
|
Recommender systems - models, methods, experiments
Peška, Ladislav ; Vojtáš, Peter (advisor) ; Jannach, Dietmar (referee) ; Krátký, Michal (referee)
This thesis investigates the area of preference learning and recommender systems. We concentrated recommending on small e-commerce vendors and efficient usage of implicit feedback. In contrast to the most published studies, we focused on investigating multiple diverse implicit indicators of user preference and substantial part of the thesis aims on defining implicit feedback, models of its combination and aggregation and also algorithms employing them in preference learning and recommending tasks. Furthermore, a part of the thesis focuses on other challenges of deploying recommender systems on small e-commerce vendors such as which recommending algorithms should be used or how to employ third party data in order to improve recommendations. The proposed models, methods and algorithms were evaluated in both off-line and on-line experiments on real world datasets and on real e-commerce vendors respectively. Datasets are included to the thesis for the sake of validation and further research. Powered by TCPDF (www.tcpdf.org)
|
|
|
Content based Recommendation from Explicit Ratings
Ferenc, Matej ; Vojtáš, Peter (advisor) ; Peška, Ladislav (referee)
In the thesis we compare several models for prediction of user preferences. The focus is mainly on Content Based models which work with metadata about objects that are recommended. These models are compared with other models which do not use metadata for recommendation. We use three datasets and three metrics to get the results of recommendation. The goal of the thesis is to find out how can the metadata about the users and the objects enhance the standard recommender models. However, the result is that the metadata can enhance recommendation in some cases, but it varies by used metrics and dataset. This enhancement is not significant.
|
guest ::
