Artsіbasov V. Methods and tools for building recommendation systems for e-commerce tasks.

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0415U004463

Thesis Registration Form

0415U004463.pdf

Applicant for

Specialization

  • 05.13.06 - Інформаційні технології

04-09-2015

Specialized Academic Board

Д 35.052.14

Lviv Polytechnic National University

Essay

This thesis addressed the task of developing scientific methods and construction tools of recommendation systems for e-commerce tasks. An improved method for calculating the coefficients of similarity vectors for the method of weighted sum. A method for providing advice to the new user that uses the association rules and does not require registration of new users to the system. The method of introduction of recommender systems profile of the new user and a new subject on the basis of categorical clustering. It has been further developed a hybrid method of predicting the recommendations that uses the fuzzy clustering с-means and the method of weighted sum.

Read more

Files

Artsibasov_ aref.doc

Artsibasov_dis.doc

Similar theses

0424U000107

Oleksandra V. Kovyrova

Models and instrumental tools of express diagnostics for using in the biology and medicine

0524U000111

Tetiana A. Honcharenko

Methodological foundations of the formation of a unified information environment for the automation of object-spatial systems in construction projects

0524U000108

Ihor M. Liakh

Methodological foundations of information technology for gene expression data processing and its application in the field of bioinformatics

0524U000074

Andrii Shyshatskyi

Intelligent methods of managing interference protection of radio communication systems under conditions of destabilizing influences

0524U000069

Gnatchuk Elizaveta Gennadievna

Theoretical and applied principles of information technology for supporting medical decision-making considering the civil law grounds