mdh.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Språkteknologiförstärkt matchning för att identifiera liknande produkter
Mälardalen University, School of Innovation, Design and Engineering.
2017 (Swedish)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesisAlternative title
NLP enhanced search for identification of similar products (English)
Abstract [en]

In this study the quality of product searching for e-commerce was examined to see how it could be improved. The objective with this thesis was to make it easier for customers to find relevant articles for the searched product. Two search engines were implemented, the first engine was built on parts from natural language processing and the second one was based on a traditional version built on keyword matching. Customers are typically presented with four different product alternatives with the product they searched for. The search results from both implementations was also evaluated with surveys from 20 individuals where they proposed how the products should be ranked. The evaluation showed that the implementation with natural language processing was more efficient in comparison with the keyword based implementation in a chosen field (kids clothing). The engine with natural language processing showed 75 % (for both genders) matching with the top four favoured products from the survey. The keyword based resulted only in 17 %. Natural language based search resulted in 83 % matching for females only. The result proves that natural language processing version is more effective in the given domain (kids clothing, 2000 product in the database) and should be in consideration when implementing search functions for e-commerce.

Abstract [sv]

I denna studie undersöktes hur kvalitén på produktsökningen i e-handeln kan förbättras. Syftet med arbetet var att underlätta för kunder att hitta relevanta artiklar till den sökta produkten. Det skapades två sökmotorer, den första en implementation med valda delar från språkteknologin och den andra en traditionell variant byggt på nyckelordsmatchning. Vanligtvis brukar besökare på en e-handelsplats erbjudas fyra alternativ tillsammans med produkten de tittar på. För att även mäta resultatet mot användare fick 20 personer rangordna förslagen. Det visade sig att implementationen med språkteknologin var avsevärt bättre i jämförelse med nyckelordssökningen inom det valda området (barnkläder). Den språkteknologibaserade sökningen överensstämde till 75 % (för båda könen) med de fyra produkterna som respondenterna ansåg vara de fyra bästa förslagen. Den traditionella nyckelordssökningen överensstämde endast till 17 %. För kvinnor stämde resultatet med 83 % för den språkteknologibaserade sökningen. Resultatet visar att en språkteknologibaserad sökning ger klart bättre resultat än en nyckelordsmatchning i den givna domänen (barnkläder, 2000 produkter i registret) och bör övervägas när det implementeras matchningsfunktioner i e-handelssystem.

Place, publisher, year, edition, pages
2017. , p. 74
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:mdh:diva-34920OAI: oai:DiVA.org:mdh-34920DiVA, id: diva2:1075714
External cooperation
POCADA AB
Subject / course
Computer Science
Supervisors
Examiners
Available from: 2017-02-21 Created: 2017-02-20 Last updated: 2017-02-21Bibliographically approved

Open Access in DiVA

fulltext(4470 kB)49 downloads
File information
File name FULLTEXT01.pdfFile size 4470 kBChecksum SHA-512
9a30bd8e04a1009af513b48ad68b8fa7828bf2800b794acb9cc7a2233102e5afc284bbde0922038a4aff298150d6ff4a712d5c509b3c83e1657a2aeb051f12db
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Bajalan, Amanj
By organisation
School of Innovation, Design and Engineering
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 49 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 151 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf