Application of a semantic product matching mechanism in open tendering e-marketplaces
Date | 05 March 2018 |
Pages | 14-30 |
Published date | 05 March 2018 |
DOI | https://doi.org/10.1108/JOPP-03-2018-002 |
Author | Ahmad Mehrbod,Aneesh Zutshi,António Grilo,Ricardo Jardim-Gonsalves |
Subject Matter | Public policy & environmental management,Politics,Public adminstration & management,Government,Economics,Public finance/economics,Taxation/public revenue |
Application of a semantic product
matching mechanism in open
tendering e-marketplaces
Ahmad Mehrbod,Aneesh Zutshi and Ant
onio Grilo
UNIDEMI, Faculdade de Ciências e Tecnologia da, Universidade Nova de Lisboa,
Campus de Caparica, Monte de Caparica, Portugal, and
Ricardo Jardim-Gonsalves
UNINOVA, Faculdade de Ciências e Tecnologia da, Universidade Nova de Lisboa,
Campus de Caparica, Monte de Caparica, Portugal
Abstract
Purpose –Searchingthe tender notices that publish every day in open tenderingwebsites is a common way
for finding business opportunity in public procurement. The heterogeneity of tender notices from various
tenderingmarketplaces is a challenge for exploiting semantictechnologies in the tender search.
Design/methodology/approach –Most of the semantic matching approaches require the data to be
structured and integratedaccording to a data model. But the integration process can be expensiveand time-
consumingespecially for multi-source data integration.
Findings –In this paper,a product search mechanism thathad been developed in an e-procurementplatform
for matching producte-catalogues is applied to the tendersearch problem. The search performance has been
comparedusing two procurement vocabularieson searching tendernotices from two major tender resources.
Originality/value –The test results show that the matching mechanism is able to find tender notices from
heterogeneous resources and different classification systems without transforming the tenders to a uniform data model.
Keywords Public procurement, Tender notice, Semantic matching, Classification systems
Paper type Research paper
Introduction
To prevent corruptionand give equal opportunities to all competitors in public procurement,
public tenders have to be published openly in many countries. The tender notices that
publish every day in procurement journals and e-marketplaces not only provides better
value for money for governments, but also act as a valuable resource for several suppliers
for finding business opportunities(Graux et al., 2012). e-Procurement (Ghimire et al.,2013)is
considered to be a strategic tool for improving the competitivenessof organizations. B2B e-
procurement chain consists of several necessary stepsincluding e-Sourcing, e-Noticing and
e-Tendering, e-Awarding and e-Contract (also called e-Reverse Auctioning), e-Ordering, e-
Invoicing and e-payment that can be summarized into two main phases: pre-award phase
(tendering) and post-awardphase (purchasing) (Kajan et al.,2012;Grilo et al., 2013).
In e-procurement, the tendering phase is usually done using reverse auction in which
suppliers bid on the services or goods that buyers need. In this procedure, different bidders
generate competing offers on tenders and look to obtain an award of business activity in
The authors would like to acknowledge FCT/MCTES the financial support of UNIDEMI through the
Strategic Project Pest-OE/EME/UI0667/2014.
JOPP
18,1
14
Journalof Public Procurement
Vol.18 No. 1, 2018
pp. 14-30
© Emerald Publishing Limited
1535-0118
DOI 10.1108/JOPP-03-2018-002
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1535-0118.htm
works, supply, or service contract (Raffa and Esposito, 2006;Du, 2009). To have a great
improvement on the accessibility and transparency of the tenders and also to provide equal
business opportunities to all suppliers, public e-Tendering Marketplaces should be
accessible publicly in the EU as well as many other countries (European Parliament, 2014;
Vaidya et al.,2006). The publicly noticed tenders in such marketplaces are used by various
companies to find businessopportunities (Grilo and Jardim-Goncalves, 2013).
Internet-based information systems and platforms are used in e-procurement to replace
one, some or all stagesof the traditional procurement process. In otherwords, e-procurement
systems may provide an end-to-end solution that covers all procurement phases or
digitalizes some importantaspects of the procurement process, such as search andselection;
with a specific emphasis on efficiency,transparency and policy in the public sector (Roman,
2013). Searching and selecting the best suitable opportunities among several published
tender calls especiallyfrom various tendering resources is a crucial yet time-consumingtask
for several businessactors in e-procurement marketplaces (Svidronovaand Mikus, 2015).
Most tendering websitesprovide keyword-based search and category-based notifications
to the subscribers for Tendersand Contract Awards. The idea behind such search system is
that any company will receive all published tenders in their business section. Accordingly,
the tender notification systems deliver tender opportunities to the company, dramatically
reducing the amount of time spent looking for these tenders.But according to a potentially
wide range of tender calls in a business sector, a supplier may receive an extensive list of
notifications from each subscribed tendering marketplace. Thismakes it difficult to find the
best-matched opportunitieswith the supplier’s product portfolio.
This problem can be solved using a search mechanism that is able to rank and sort the
tender calls coming from variousresources based on their similarity to a supplier’s products
or services. Consequently, suppliers can save time in searching and work on preparing
proposals for the mostsimilar calls to their products that definitely have more chanceto win
the competition. To reach this goal, the product information provided by a supplier can be
used by a product search mechanism to find and recommend similar product requests
(Julashokri et al.,2011) and tender calls.
The purpose of this paper is to use product data provided in tenders by buyers and
match it with the product data provided in e-catalogues by the suppliers to solve the
opportunity search problemin e-procurement marketplaces. Therefore, reusing of a product
matching mechanism is proposed to solve the opportunity finding problem. The following
subsections study the product matching scenarios that the search mechanism has to cover
and the problems which the matchingmechanism has to encounter to be used in opportunity
search in e-procurementmarketplaces.
Matching scenario
One of the main factors that make the procurement process and consequently the
procurement marketplaces different from the wholesale process and marketplaces such as
Alibaba is the direction of the purchasing process. In wholesale websites the producers,
suppliers and distributors start the process by offering the goods on the website; and the
buyers search to find their needed goods, inquiry the price and features, negotiate the
purchasing, payment and delivery process and then order. But in a procurement
marketplace, the purchasing process starts by the buying organization. This process is
called Reverse Auction (in comparison with the forward auction) and is used in B2B
procurement to obtain the best priceby encouraging competition between the suppliers (Jap,
2007). In pre-award phase, the buying organization makes a call for tender based on its
requirements, publishesit publicly or sends it to a prequalified list of suppliers based on the
Semantic
product
matching
mechanism
15
To continue reading
Request your trial