Priporočilni sistem za spletno trgovino

diplomsko delo

Tomaž Silič (Author), Igor Kononenko (Mentor)

Abstract

Zaradi vse večjih potreb po hitri obdelavi velike količine podatkov se v poslovnem svetu vedno bolj pogosto uporabljajo metode podatkovnega rudarjenja znotraj različnih vrst sistemov CRM. Sistemi CRM se dopolnjujejo s tako imenovanimi priporočilnimi sistemi, ki tako strankam kot prodajalcem priporočajo izbiro raznih akcij znotraj sistema. V diplomskem delu smo pregledali obstoječe tehnike priporočilnih sistemov in posodobili manjši sistem CRM oz. spletno trgovino s priporočilnim sistemom za stranke in prodajalce. Za prodajalce smo s pomočjo algoritma Apriori tvorili povezovalna pravila med artikli glede na predhodna naročila kupcev. Povezovalna pravila so se izkazala za neuporabna, saj med artikli spletne trgovine ni bilo tesnih povezav. Za stranke pa smo po algoritmu ID3 zgradili drevo za priporočanje artiklov, ki bi utegnili stranke zanimati poleg že ogledanih artiklov. Zgradili smo dve drevesi na osnovi zgodovine obiska spletne trgovine. Podatke o obisku smo pridobili iz sistema Google Analytics.

Keywords

CRM;spletna trgovina;priporočilni sistemi;podatkovno rudarjenje;računalništvo;univerzitetni študij;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2015
Typology:	2.11 - Undergraduate Thesis
Organization:	UL FRI - Faculty of Computer and Information Science
Publisher:	[T. Silič]
UDC:	004.85:004.738.5:339 (043.2)
COBISS:	1536615363
Views:	1943
Downloads:	375
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Recommender system for a web store
Secondary abstract:	Due to the increasing demand for high-speed processing of large amounts of data in the business world, data mining is becoming widely used within the different types of CRM systems. CRM systems are complemented with recommender systems that both customers and salespeople use for selecting various actions within the system. In this thesis we reviewed the existing techniques for recommendation systems and updated the CRM system i.e. an online store with a recommendation system for customers and salespeople. For salespeople we are using an algorithm Apriori which formed the association rules between products compared to previous customers' orders. Association rules have proved to be useless, because there is no close links between products. For customers, we are using ID3 algorithm to built a recommendation tree to recommend products which may be of interest to them. We have built two trees based on the history of online shop visits. Visits data were obtained from Google Analytics system.
Secondary keywords:	CRM;web shop;recommendation system;data mining;computer science;diploma;
File type:	application/pdf
Type (COBISS):	Undergraduate thesis
Study programme:	1000475
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages:	36 str.
ID:	9057320