2026-07-09T10:01:50.198Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:1:8233 2021-05-18T21:47:56Z 1

Optimization of e-Commerce search engine with approximate string matching technique 2018 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/1/8233 M. Dobrojević T. Golubović An important factor in the success of any web shop is ability to provide relevant search results to customers. Each web shop operator has its distinctive way of naming and grouping products into categories or site sections. Subtle variations in search queries, although easily recognized by humans, may present a significant obstacle for e-commerce search engine in order to provide relevant search results. Big data based recommendations heavily rely on mathematical approach, demand highly trained analysts and loose any real personalization in the process. In such scenarios approximate string matching technique, or better known as fuzzy search, may prove to be useful. Although under certain scenarios fuzzy search may produce wrong search results, simple corrections in the search logic are sufficient enough to overcome this problem. On the other hand, multiple string matching iterations bring flexibility to the process, simplifying results testing and verification during model building phase. conferenceObject 116 121 7th International Symposium on Industrial Engineering Proceedings