COMMON TEXTILE VOCABULARIES AND DOCUMENTS

A Conceptual Foundation of a Globally Interoperable Textile e-Marketplace

Jingzhi Guo and Zhuo Hu

Department of Computer and Information Science, University of Macau, Av. Padre Tomás, Pereira, S.J., Taipa, Macau

Keywords: Textile electronic marketplace, business vocabulary, business document, product data integration, vocabu-

lary integration, document integration, semantic consistency maintenance, semantic interoperability, elec-

tronic commerce, electronic business.

Abstract: This paper has proposed a novel common textile vocabulary and document framework (TexVDF) in a col-

laborative network to enable cross-domain level business information sharing and business document ex-

change in a semantically consistent way. The approach to this framework is motivated through presenting

some real-world examples of business inquiries with product specifications. By these examples, two prob-

lems are detected on how to achieve semantic commonality between cross-domain level business vocabular-

ies for textile e-Marketplace mediators and how to allow specificity of cross-domain level common business

document templates for local textile e-Marketplace mediators yet still maintaining semantic consistency. To

solve these two problems, this paper has firstly reviewed CONEX technologies relevant to the newly devel-

oped TexVDF approach, which includes a TexVDF framework, a P2P collaborative textile concept mapping

model and a textile business document template model. These two models have been demonstrated by ex-

amples to see how they should work.

1 INTRODUCTION

In global textile market, many textile firms face

great challenges of increasing global competition to

cope with quicker responsiveness of competitors

(Teng et al. 2006), better global partnership, and

advancing technology for manufacturing and dy-

namic consumer demand (Ostic 1997). To meet

these challenges, textile industry needs to build

global e-Marketplaces, where producer, supplier,

and retailers can efficiently communicate and ex-

change information to strengthen their competitive-

ness. However, textile industry is complex. It in-

volves raw materials such as cotton, silk and poly-

mer that produce both natural and synthetic fibres,

which again are converted into many kinds of fab-

rics and finally to become products such as carpet

and apparel.

Besides, textile industry has thousands of large,

medium and small sized retailers and manufactures

(Teng et al. 2006). They vigorously trade with each

other. This indicates that a global e-Marketplace in

design must be able to manage the flow of millions

of textile products between a very large number of

firms. This requires the information exchanged by

computers be understandable by all textile partici-

pants.

Making textile information understandable be-

tween participants on e-Marketplace is an important

information interoperability issue (Guo 2007). It

requires building a globally interoperable textile e-

Marketplace by integrating heterogeneous textile

information systems of all participated firms. This at

least involves two aspects: the integration of busi-

ness vocabulary used by all textile firms and the

integration of exchanged business documents such

as inquiries, offers, counteroffers and orders.

This paper aims to propose a novel common tex-

tile vocabulary and document framework (TexVDF)

to semantically integrate complex textile firms to

enable them to participate in textile e-Marketplace,

and also to facilitate the proposed approach as the

foundation of the future design of globally interop-

erable textile e-Marketplace.

The rest of the paper will be arranged as follows:

Section 2 will provide a motivational scenario to

raise our discussion issues. Section 3 will briefly

introduce the relevant technologies and propose a

new TexVDF approach to lay a solid conceptual

foundation for future design of textile e-

Marketplace. Section 4 will exemplify TexVDF

approach. Related work is discussed in Section 5.

Finally, the conclusion with a contribution list of this

paper is given, together with the required future

work.

Guo J. and Hu Z. (2008).

COMMON TEXTILE VOCABULARIES AND DOCUMENTS - A Conceptual Foundation of a Globally Interoperable Textile e-Marketplace.

In Proceedings of the International Conference on e-Business, pages 81-88

DOI: 10.5220/0001909600810088

 SciTePress

2 A MOTIVATIONAL SCENARIO

e-Marketplace has four properties: distribution,

autonomy, interdependence and emergence (Guo

2007). These also apply to textile e-Marketplaces.

However, textile e-Marketplaces have a more spe-

cific property of levelled-domain interdependence,

i.e. one level of textile e-Marketplace vocabularies

and documents are tightly-coupled with another

level of textile e-Marketplace vocabularies and

documents. Since they belong to different domains,

they have specific requirements, For example, they

involve interdependence between levels of industries

of raw supplies (e.g. cotton, silk and polymer), fibre,

yarn, thread, fabric, printing and dyeing, and gar-

ment.

To motivate the problem to solve, Table 1 to 5

summarize the inquiries created by different levels

of textile industries to show their interdependence.

Table 1: Raw Cotton Inquiry from Material Industry.

Concept Concept Value

Product name Indian raw cotton

Type Shankar-6

Staple length 29 mm OR 28.5 mm

Strength 29 gpt or 28 GPT minimum

MIC 3.5-4.9

Table 2: Polyester Fibre Inquiry from Fibre Industry.

Concept Concept Value

Name Polyester Stable Fibre

Colour Semi-dull raw white

Actual Denier 1.40 + / -0.05 (DE)

Dry Tenacity 6.50 + / -0.05 (G/D)

Dry Elongation 30. + / -1.5

Crimp number 13.00 + / -3.00 EA / Inch

Degree of Crimp 13.00 + / -0.39

Shrinkage 7 + / -0.07

Table 3: Blended Yarn Inquiry from Yarn Industry.

Concept Concept Value

Name Yarn

Colour white

Type Carded

Composition

Cotton 60%

Polyester 40%

Count 32 s

Quality high

Application Quality knitting and weaving fabrics

An analysis to the above Tables reveals at least

two of the following problems:

(1) Term usages are not only industry domain-

specific. The higher-level industry has a high fre-

quency to use the vocabularies of lower-level indus-

try, but practically each level of industry has its own

explanations on their terms in vocabularies.

This problem can be abstracted as a research is-

sue of cross-domain level business vocabulary

commonality, which states that all levels of textile

industry shall be able to semantically communicate

with each other via a common vocabulary model.

Table 4: Dyed Fabric Inquiry from Fabric Industry.

Concept Concept Value

Name Fabric

Type Combed

Colour Yarn dyed

Technique woven

Style Jacquard

Usage Garment

Composition

Cotton 60%

Polyester 40%

Construction

45 s

Density

130

Width 57/58”

Table 5: Men’s Coat Inquiry from Garment Industry.

Concept Concept Value

Name Coat

Fabric

Cotton 60%, ring, spun

Polyester 60%

Weight 300 grams/square meter

Specification

Zipper Front

Pocket Pouch

Cuff rib, lycra (cotton 60%, polyester 40%)

Bottom rib, lycra (cotton 60%, polyester 40%)

Size S, M, L, LL, LLL, LLLL

Drawcord Fabric

Use Men

This problem can be abstracted as a research is-

sue of cross-domain level business vocabulary

commonality, which states that all levels of textile

industry shall be able to semantically communicate

with each other via a common vocabulary model.

(2) Inquiry templates, as shown in Tables, from

different levels of industry are different in syntactic

forms and semantic use of terms, though they are all

called as inquiry sheets. This phenomenon implies

that a same type of business documents must be

treated differently in specific domain-level industry.

We abstract this phenomenon as a research issue

of cross-domain level common business document

specificity, which states that all domain-levels of

textile industry shall be able to personalize docu-

ment templates from common document templates.

3 TEXVDF APPROACH

In this section, we will propose a novel common

Textile Vocabulary and Document Framework

(TexVDF) to solve the above two problems to lay a

ICE-B 2008 - International Conference on e-Business

solid foundation for future globally interoperable

textile e-Marketplace. We vision that this e-

Marketplace will be a common textile information

space that all kinds of services and collaborative

activities will be enabled, based on our designed

TexVDF solution.

In the following, we will first discuss the rele-

vant technologies and then propose the solution

framework and its details.

3.1 Relevant Technologies

In CONEX research (Guo 2008), a generic e-

Marketplace is designed like a multi-sons solar sys-

tem as a set of common collaborative service media-

tors, each having a set of local collaborative service

designers for their own service users on CONEX

Network (ConexNet). It can be shown in Figure 1.

Figure 1: CONEX E-Marketplace.

In this CONEX e-Marketplace, business vocabu-

laries and documents are firstly collaboratively de-

signed at mediators (M) on a P2P collaborative net-

work. Mediators include designer roles of common

vocabulary design, dictionary design and document

template design. These designers are knowledge

experts (or knowledge engineers) and can make sure

that things go right. Thus, they are called dominators

(D) when their results are sold to local firms. The

local firms subscribing or purchasing mediators’

design services localizes these services into their

own personalized forms, that is, local business vo-

cabularies and document templates. The local de-

signers are knowledge workers but are not experts.

Thus, they are followers (F) of the dominators (D).

The collaboration between dominators and followers

is a dominator-to-follower (D2F) relationship on a

point-to-point communication network. The follower

cannot modify the common design but follow to

generate their own. When local designers have de-

signed local information about firm-based vocabu-

lary and document templates, their users of their

own firms can then automatically exchange business

information, such as inquiries and offers.

In CONEX e-Marketplace, vocabularies and

documents are designed following Product Map (PM)

theory (Guo 2008). Its implementation is specified

in CONEX Grammar (Guo 2008).

The information exchange by PM from one local

firm (L) to another local firm (L) follows a concept

supply chain (Guo 2008), such that:

Concept(L1) map onto Concept(M1) map onto

Concept(M2) map onto Concept(L2),

where concept(L1) of L1 finally arrive at L2 as

concept(L2).

Since concepts of L1, M1, M2 and L2 are all col-

laboratively created and mapped, there is no seman-

tic inconsistency between them in theory. Thus,

CONEX e-Marketplace provides an approach of

accurate yet automatic information exchange in the

perspectives of all users (U) of L.

3.2 TexVDF Framework

The TexVDF framework follows the design of

CONEX e-Marketplace but adds the new layer

thought to the framework. To provide a smooth

discussion, we illustrate this framework in Figure 2,

where different levels of textile industry have been

separated but integrated in a coherent collaboration

framework.

In Figure 2, CONEX e-Marketplace has been ex-

tended to include six layers of textile e-

Marketplaces, which are:

• Raw material e-Marketplace, which focuses on

trading cotton, silk, wool, fur, feather, plant,

polyester, polypropylene, polyethylene, etc.

• Fibre e-Marketplace, which focuses on trading

fibres made from cotton, silk, polymer (e.g.

polyester, Dacron, nylon, Micron), plant (e.g.

corn, soybean), etc.

• Yarn and thread e-Marketplace, which focuses

on trading yarns of cotton, polyester, blended

polyester, wool, plant, different threads, etc.

• Fabric e-Marketplace, which focuses on trading

woven and non-woven fabrics of cotton, polyes-

ter, blended polyester, metallic, etc.

• Printing and Dyeing e-Marketplace, which

focuses on trading dye, dyeing services of vari-

ous yarns, threads, fabrics, printing, etc.

• Garment e-Marketplace, which focuses on fin-

ished textile products like jacket, pullover, shirt,

protective garment, gown, gloves, etc.

By this layered classification of textile e-

Marketplaces, a new technology of P2P collabora-

tive mapping between e-Marketplace mediators is

developed to solve the problems stated in Section 2.

This technology complements the previous CONEX

technology to enable cross-domain level vocabulary

and document interoperability.

M M

P2P Collaboration

D2F Collaboration D2F Collaboration

Use

U: User L: Local Designer M: Mediator

COMMON TEXTILE VOCABULARIES AND DOCUMENTS - A Conceptual Foundation of a Globally Interoperable

Textile e-Marketplace

Figure 2: TexVDF Framework.

A standard terminology servicing centre is added

to enable the adoption of existing standard terms

such as standard measure, currency code, and de

facto standard use of textile terms. This service will

reduce the collaborative work effort occurred in both

vocabulary and document templates for cross do-

main levelled textile e-Marketplaces.

Particularly, the TexVDF framework adopts a

collaborative mapping solution to achieve semantic

commonality of cross-domain level vocabularies and

a specificity solution to using common textile docu-

ment template.

In the next two subsections, we will elaborate the

design of these two solutions.

3.3 P2P Collaborative Mapping

A P2P collaborative mapping model (CMM) is a

solution to enabling various mediators of levelled

textile e-Marketplaces to collaboratively work to-

gether to map their textile concepts in a common

level but with possible expression variations for their

own. This model is necessary because there are

many synonyms and homonyms appeared in textile

industry. Each textile mediator has its own concept

mediation context and is not possible to define all

synonyms and homonyms for each textile concept.

The impossibility for each textile mediator to in-

clude all is due to the complexity that textile terms

are developed from different contexts, where syno-

nyms and homonyms come from various textile

types, usages of dated and obsolete terms, term deri-

vations, languages of origin, and toponyms (Leech

1999).

For example, “baby combing wool” has variation

of “French combing wool”; “linen lawn” is synony-

mous with “handkerchief linen”; and “drab” is a

homonym with meanings of “colour of a moderate

to grayish or light grayish yellowish brown or light

olive brown” and “cloth of a light dull brown or

grayish brown or unbleached natural colour, espe-

cially a heavy woollen or cotton fabric”.

To bind the concept expressions of individual

mediators with a common concept expression, the

mediators’ concept expressions must be framed in a

common structure, where they can freely express

their ideas of categorizing their individual textile

concepts. CCM provides this function in the follow-

ing form, such that:

Definition 1 (CMM). Collaborative Mapping

Model

Given a set of individual concept expressions σ

, ..., σ

and uniquely identified by iid

, iid

, ..., iid

under different mediators’ contexts x

, x

, ..., x

about a piece of common formal concept FC, de-

fined by AN, uniquely identified by IID, typed by

CT, inherited from IF, concept or context of CO, and

possibly referenced to (RT) a standard term of a

terminology, such that:

Γ |- concept[IID, FC, AN, CT, IF, CO, RT] ⇒

map[concept[IID, iid

, σ

)@x

], ..., concept[IID,

iid

, σ

]@x

where, “Γ” is a P2P collaborative mapping en-

gine that enables all mediators to work together in

real-time, “|-”is a declaration that all mapping follow

the mapping structure that is defined, and “⇒” refers

to that the semantics of each mediator’s concept

strictly follows the semantics of common concept

for all.

The above CMM has following principles:

Raw Material

Mediators

Fibre

Mediators

Yarn & Thread

Mediators

Fabric

Mediators

Printing &

Dying

Garment

Mediators

P2P Colla-

borative

Mapping

Network

P2P Collabo-

ration Net-

work

Standar

Terminolo-

Servicing

Local

Designer

Local

Designer

Local

Designer

Local

Designer

Local

Designer

Local

Designer

Local

User

Local

User

Local

User

Local

User

Local

User

Local

User

use

D2F

use

Raw Material

E-Marketplace

Fibre

E-Marketplace

Yarn & Thread

E-Marketplace

Fabric

E-Marketplace

Printing &

Dying

Garment

E-Marketplace

Vocabularies

Document Templates

ICE-B 2008 - International Conference on e-Business

(1) Each mediator’s semantics about the same con-

cept must be semantically consistent by P2P

collaboration following CMM structure.

(2) Personalized concept expressions of individual

mediators, rather than the formal concept FC,

are allowed to adapt to the local mediator’s e-

Marketplace environment but have to be

mapped onto common concept via IID.

(3) Standard terms of terminology, from standard

terminology servicing centre shown in Figure 1,

are encouraged to be referenced during common

concept design.

Particular to CMM, some details are important to

mention.

Definitional annotation AN. It is a full definition

of a common concept, and is not a single word or

phrase that may not fully capture the meaning of the

concept, or may lead to sense ambiguity.

Formal concept FC. It is a machine-readable

term about a common concept. It may capture full or

almost full meaning of the concept definition AN,

but not guaranteed for accuracy. It is primarily used

for information retrieval for search services. FC is

typed as a set to include multiple words and phrases

with exact or similar meanings to the defined con-

cept such as abbreviation.

Internal unique concept identifier IID. It is

unique identifier of AN with semantic causal order

relationship such as AN ⇒ IID. It implies that any

IID cannot be created to use without the meaning of

AN conveyed in IID. If such happens, it will be

prohibited to process.

Reference to a concept RT. It is a reference to a

semantically equivalent concept, often a well-

defined term in terminology or an already-defined

vocabulary by CMM. It is typed as a namespace use.

Context of a higher level concept CO. It defines

the direct context of the current concept in a vocabu-

lary hierarchy. For example, given “domestic appli-

ances (domestic refrigerators)”, the “domestic appli-

ances” is the direct context of “domestic refrigera-

tors”. It is similar to a broad term (BT) in relation to

a narrow term (NT) in controlled vocabulary (Fidel

1999). The CO is important for efficient and accu-

rate information exchange. It is also very useful for

accurate machine translation by word sense disam-

biguation (Vickrey et al. 2005) through CO context.

Inheritance from sources IF. It defines that

where the concept is inherited or derived. It states

the origin of the concept. It is useful for improving

information retrieval and disambiguating the sense

of the concept when the concept is applied in ma-

chine translation. For example, “ramie cotton

blended fabric” is inherited from both “ramie” and

“cotton” and under the context_of (CO) of neither

“ramie fabric” nor “cotton fabric” but “blended

fabric”. The inherited terms of “ramie” and “cotton”

are useful to infer the concept meaning.

Concept type CT. It defines in which term type

the concept applies. For example, “bombazine”

means “a twilled or corded dress-material, composed

of silk and worsted; sometimes also of cotton and

worsted, or of worsted alone”. Thus, it belongs to

the domains of “silk” and “cotton” as a noun-form,

but it also belongs to the large domain of “textile” as

an adjective form to refer to “worsted”.

With above descriptions, the common concept

concept[IID, FC, AN, CT, IF, CO, RT] is clear for

individual mediators to collaboratively map onto

their own local forms in map[concept[IID, iid

)@x

], ..., concept[IID, iid

, σ

]@x

]. In the local

form, practically, the context x

can be designed as

an individual mediator’s unique identifier (LID) to

refer to its specific context.

3.4 Document Template Model

A textile document template model (Doclate) is a

solution to enabling specifying business document

templates to a domain-specific level but still being

able to utilize the common document templates

prepared in advance and to consistently use common

concepts designed by CMM.

To realize this model, this paper regards a busi-

ness document template a set of hierarchical con-

cepts and improves the previous work of CODEX

(Guo 2006) by removing P2P collaboration require-

ment and adding RT to CMM, such that:

Definition 2 (Doclate). Document Template Model

Given a set of concepts defined in CMM, there is

a common document template (com) with a set of

concepts (called elemon) hierarchically identified as

IID, annotated by AN, in the context of CO, pre-

sented as DP, referenced to RT of CMM, and have

occurrence OC; each “element” has a reification

structure “value” to reify the concept with represen-

tation format PT, data type DT and possible function

FN for managing reification. This “com” will further

be extended to (⇒) “loc” to provide specificity of

“com”, such that:

Γ |- com: elemon[IID, AN, CO, DP, OC,

RT](value[PT, DT, FN]) ⇒ loc: elemon[IID, AN,

CO, DP, OC, RT](value[PT, DT, FN]),

where “Γ” is a business document template edit-

ing engine, “|-” declares that the editing follows the

structure defined.

The above model has the following principles:

(1) Each common document template “com” is only

a semi-finished template and could be further

COMMON TEXTILE VOCABULARIES AND DOCUMENTS - A Conceptual Foundation of a Globally Interoperable

Textile e-Marketplace

specified in any way to “loc”, but both must fol-

low the Doclate structure model.

(2) Both “com” and “loc” can only create “elemon”

concepts through RT to CMM vocabularies.

Particularly, the Doclate structure elements have the

following semantics:

AN, IID, CO and RT. They exactly have the same

interpretation as in CMM, but RT refer to CMM.

Document concept display phrase DP. It is a

phase for visual display to represent concept. For

example, a concept of “an appliance, a cabinet, or a

room for storing food or other substances at a low

temperature” could be visually displayed as “refrig-

erator” or “domestic refrigerators” as needed in the

different designs of Doclate template.

Concept occurrence OC. It defines occurrence of

concept in designed Doclate template. The concept

occurrence may happen. For example, “product item”

in a purchasing order may occur many times for

different purchased items.

Value concept VALUE. This is a reification sym-

bol to introduce a reification of a concept to a par-

ticular concept, for example, “colour” → “red”.

Presentation style of reified concept PT. It de-

fines how a reified concept should be displayed. For

example, “1” could be displayed as “1”, “one”, or

“USD1/piece”. It is a logic module and implemented

in a remote namespace.

Data type of reified concept DT. It defines the

data type of the reified concept, for example, “string”

or “decimal”.

Operational function of reified concept FN. It de-

fines how the reified concept value could be com-

puted, for example, automatically generated reified

“date” value, reified computational group concept

“total”, or a result of a logic module.

With the above descriptions, common document

templates could be easily specified locally but accu-

rately maintain semantic consistency between using

parties of different levels of textile e-Marketplace.

Table 6: Textile Common Vocabulary for Garment.

concept[iid=“1.1” fc=“cotton” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“1.2” fc=“polyester” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“2” fc=“fabric” an=“-” ct=“n” if=“-” co=“-” rt=“-”]

concept[iid=“2.1” fc=“rib” an=“knit ribbing” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“3” fc=“weight” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4” fc=“coat” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4.1” fc=“zipper” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4.2” fc=“pocket” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4.3” fc=“hood” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4.3.1” fc=“drawcord” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4.4” fc=“cuff” an=“-” ct=“-” if=“-” co=“-” rt=“-”]

concept[iid=“4.5” fc=“bottom” an=“-” ct=“garment” if=“-” co=“-” rt=“-”]

concept[iid=“5” fc=“size” an=“-” ct=“-” if=“-” co=“-” rt=“msr:123456”]

concept[iid=“6” fc=“use for” an=“-” ct=“-” if=“-” co=“-” rt=“hba:12345”]

concept[iid=“7” fc=“specification” an=“-” ct=“-” if=“-” co=“-” rt=“spc:1”]

4 COAT EXAMPLE ON TEXVDF

In this section, we demonstrate TexVDF approach

through a garment inquiry specified in Table 5. The

following examples will adopt the written form of

vector concept tree like “1.i...i” to represent the

concept hierarchy of any vocabulary and document

template (Guo 2008).

4.1 CMM for Coat Vocabularies

To implement the garment inquiry example, we first

design the common vocabulary, shown in Table 6,

based on CMM of definition 1.

With the above collaboratively designed garment

vocabulary commonly for all levels of textile e-

Marketplace mediators, individual mediators can

localize their personalized vocabularies applicable to

their own e-Marketplaces. Table 7 shows the local-

ized mapping for one English mediator x1 and a

Chinese mediator x2 for their own e-Marketplaces.

Table 7: Mapping of Local Vocabulary onto Common

Vocabulary.

map[iid=“1.1”, (iid1=“aa” an=“cotton”)x1, (iid2=“111” an=“棉花”)x2]

map[iid=“1.2”, (iid1=“ab” an=“polyester”)x1, (iid2=“112” an=“聚酯”)x2]

map[iid=“2”, (iid1=“b” an=“fabric”)x1, (iid2=“22” an=“面料”)x2]

map[iid=“2.1”, (iid1=“ba” an=“ribbing”)x1, (iid2=“221” an=“针织布”)x2]

map[iid=“3”, (iid1=“c” an=“weight”)x1, (iid2=“33” an=“重量”)x2]

map[iid=“4”, (iid1=“d” an=“coat”)x1, (iid2=“44” an=“上衣”)x2]

map[iid=“4.1”, (iid1=“da” an=“zipper”)x1, (iid2=“441” an=“拉链”)x2]

map[iid=“4.2”, (iid1=“db” an=“pocket”)x1, (iid2=“442” an=“口袋”)x2]

map[iid=“4.3”, (iid1=“dc” an=“hood”)x1, (iid2=“443” an=“帽子”)x2]

map[iid=“4.3.1”, (iid1=“dca” an=“drawcord”)x1, (iid2=“4411” an=“拉绳”)x2]

map[iid=“4.4”, (iid1=“dd” an=“cuff”)x1, (iid2=“444” an=“袖口”)x2]

map[iid=“4.5”, (iid1=“de” an=“bottom”)x1, (iid2=“445” an=“下摆”)x2]

map[iid=“5”, (iid1=“e” an=“size”)x1, (iid2=“55” an=“尺寸”)x2]

map[iid=“6”, (iid1=“f” an=“used by”)x1, (iid2=“66” an=“适用”)x2]

map[iid=“7”, (iid1=“g” an=“cuff”)x1, (iid2=“77” an=“规格”)x2]

Given the above local mapping onto the common

vocabulary shown in Table 6, the users of local

textile e-Marketplace of x1 and x2 can then ex-

change business information without any semantic

conflicts.

4.2 Doclate for Coat Inquiry

In this part, we exemplify the work of Doclate mod-

el in the example of Table 8 and Table 9.

Table 8: Common Document Template for Inquiry.

elemon[iid=“1” an=“” co=“” dp=“coat” oc=“” rt=“4”]

elemon[iid=“1.1” an=“-” co=“” dp=“fabric” oc=“” rt=“2”]

elemon[iid=“1.2” an=“” co=“” dp=“size” oc=“” rt=“5”]

elemon[iid=“1.3” an=“” co=“” dp=“use for” oc=“” rt=“6”]

elemon[iid=“1.4” an=“” co=“” dp=“specification” oc=“” rt=“7”]

ICE-B 2008 - International Conference on e-Business

In this simple document template, some content

of the garment inquiry sheet has been designed. It

then can be further defined by local e-Marketplace

designers of document templates. The function of

common level document templates is to reduce the

time of local e-Marketplace document designers and

thus to lower the e-Marketplace cost.

When local e-Marketplace designers obtain the

rough document templates, they personalize them as

their own needs, as shown in the example Table 9.

Table 9: A Particular Garment Inquiry Sheet Template.

elemon[iid=“1” an=“” co=“” dp=“” oc=“coat” rt=“4”]

elemon[iid=“1.1” an=“-” co=“” dp=“fabric” oc=“” rt=“2”]

elemon[iid=“1.1.1” an=“” co=“” dp=“weight” oc=“” rt=“3”]

elemon[iid=“1.1.2” an=“” co=“” dp=“cotton” oc=“” rt=“1.1”]

elemon[iid=“1.1.3” an=“” co=“” dp=“polyester” oc=“” rt=“1.2”]

elemon[iid=“1.2” an=“” co=“” dp=“size” oc=“” rt=“5”]

elemon[iid=“1.3” an=“” co=“” dp=“use for” oc=“” rt=“6”]

elemon[iid=“1.4” an=“” co=“” dp=“specification” oc=“” rt=“7”]

elemon[iid=“1.4.1” an=“” co=“” dp=“zipper” oc=“” rt=“4.1”]

elemon[iid=“1.4.2” an=“” co=“” dp=“pocket” oc=“” rt=“4.2”]

elemon[iid=“1.4.3” an=“” co=“” dp=“hood” oc=“” rt=“4.3”]

elemon[iid=“1.4.3.1” an=“” co=“” dp=“drawcord” oc=“” rt=“4.3.1”]

elemon[iid=“1.4.4” an=“” co=“” dp=“cuff” oc=“” rt=“4.4”]

elemon[iid=“1.4.5” an=“” co=“” dp=“bottom” oc=“” rt=“4.5”]

elemon[iid=“1.4.1” an=“” co=“” dp=“zipper” oc=“” rt=“4.1”]

In the above Table, a more specified garment in-

quiry sheet template has been designed. With this

document template, users of local textile e-

Marketplace can reify the template by filling the

“value” information to automate inquiry exchange.

5 RELATED WORK

Textile e-Marketplace design that enables business

information sharing and business document ex-

change is an important research field. In this design,

semantic integration of textile vocabularies and

documents are the foundation for a globally interop-

erable textile e-Marketplace. Currently, active re-

searches can be found in DAMA and Moda-ML.

In U.S., TEXNET is a textile industry data-

sharing network to address data sharing among busi-

ness partners. It presents shared data in the screen or

saved it in standard formats on a local platform

(Lovejoy, a). Particular to the textile e-Marketplace,

DAMA (Chapman et al. 2000) is a project of such

type. DAMA applies a pipeline analysis method

(Lovejoy, b). In DAMA’s research, supply chain

concept is adopted through Supply Chain Integration

Program (SCIP), where an inter-enterprise decision

support tool is developed to analyze supply chain

tradeoffs. DAMA is designed to use TEXNET for

data sharing. It intends to support information shar-

ing and decision making between firms of retail,

apparel, textile and fibre within a particular supply

chain. While supply chain method for textile e-

Marketplace integration is worth investigating, it has

some entry limitations for small and medium sized

enterprises to participate in.

In Europe, MODA-M (MODA-ML; Leech 1999)

is a research on textile e-Marketplace. Its approach

is based on the exchange of standardized XML

documents, where ebXML protocol has been

adopted to transfer XML messages. Moda-ML is

ontology-based in design. It generates a modular

ontology where each basic concept can be managed

independently from the others and is identified by its

own namespace (Gessa 2007). It has defined a

common platform (Gessa et al. 2004), which at-

tempts to be adopted by firms to improve their inter-

operability. Ontology-based ebXML document ex-

change is an attractive approach and fits in most

research prototypes in many other e-Marketplaces.

However, its success depends on the ontology inter-

operability by its own in semantic level.

It is necessary to make a clear distinction be-

tween the three general strategies of integrating

heterogeneous business information (Guo 2008).

They are mandatory standardization (a standard is

enforced for all participants, in which heterogeneous

information integration between standard systems

and the participants’ local systems is the task of the

local participants), automated mediation (an intelli-

gent agent as a mediator to mediate heterogeneous

information between disparate participants’ systems

based on predefined rules, in which if no rules can

be applied, mediation of heterogeneous concepts is

not accurate), and collaborative conceptualization

(heterogeneous business concepts between disparate

local participants’ systems must be collaboratively

mapped on a higher level common system before

they can be exchanged).

This paper adopts collaborative conceptualiza-

tion strategy to avoid the weakness of mandatory

compliance of standards by users and inaccurate

concept mediation by intelligent mediator. Under

this strategy, mediator is designed as a collaborative

mediator to ensure the semantic consistency between

heterogeneous concepts. This strategy absorbs some

of the merits from both mandatory standardization

and automated mediation strategies, but it creates no

standards and mixes automated agents with human.

6 CONCLUSIONS

This paper has proposed a novel common textile

vocabulary and document framework (TexVDF) in a

collaborative network to enable cross-domain level

COMMON TEXTILE VOCABULARIES AND DOCUMENTS - A Conceptual Foundation of a Globally Interoperable

Textile e-Marketplace

business information sharing and business document

exchange in a semantically consistent way. The

approach to this framework is motivated through

presenting some real-world examples of business

inquiries with product specifications. By these ex-

amples, two problems are detected on how to

achieve semantic commonality between cross-

domain level business vocabularies for textile e-

Marketplace mediators and how to allow specificity

of cross-domain level common business document

templates for local textile e-Marketplace mediators

yet still maintaining semantic consistency. To solve

these two problems, this paper has firstly reviewed

CONEX technologies. Then, TexVDF approach is

presented in a TexVDF framework, which is an

improvement of previous CONEX model. To realize

TexVDF framework, a P2P collaborative concept

mapping model and a textile business document

template model have been developed. The former

has resolved the problem of semantic commonality

of local mediators’ individual vocabularies, and the

latter has solved the problem of designing common

business document templates and their flexible

specificity to document templates adaptable to local

textile e-Marketplaces. These two solution models

have been demonstrated in examples to see how they

could work.

The TexVDF approach has been provided as a

conceptual foundation for future design of globally

interoperable textile e-Marketplaces. It has advan-

tages compared existing solutions. (1) It does not

enforce standards on business vocabulary and

document templates. This implies a flexible solution

to semantic consistency maintenance between par-

ticipated textile e-Marketplaces. (2) The new

framework it provides allows cross-domain level

semantic interoperability but still enable personaliza-

tion. (3) Useful terminology standards are welcome

to be flexibly integrated into the new framework. It

implies a standard integration but not rigid. (4) Vo-

cabulary design between e-Marketplace mediators is

collaborative in real-time. This enables semantic

accuracy and avoids erroneous inference between

individual e-Marketplaces for those taking the ap-

proach of independent vocabulary design. (5)

Document template design and specificity adopt a

simply hierarchical document structure where each

document element concept referenced to a well-

defined collaborative concept vocabularies in CMM.

This enables simple creation and use of document

templates.

Currently, the improved structure specification of

TexCVF framework in terms of XPM is in final

release stage. The future work of this paper will be

the implementation of the textile e-Marketplace

based on this conceptual foundation.

ACKNOWLEDGEMENTS

The work reported in this paper has been supported

by University of Macau Research Grand.

REFERENCES

Censoni, P., De Sabbata, P., Cucchiara, G., Vitali, F.,

Mainetti, L. and T. Imolesi, 2002. MODA-ML: a Ver-

tical Framework for the Textile-Clothing Sector based

on XML and SOAP. In Challenges and Achievements

in E-business and E-work. IOS Press.

Chapman, L. and M. Petersen, 2000. Demand Activated

Manufacturing Architecture (DAMA) Model for Sup-

ply Chain Collaboration. In: International Conference

on Modeling and Analysis of Semiconductor Manu-

facturing.

Fidel, R., 1999. Searchers' selection of search keys: II.

Controlled vocabulary or free-text searching. Journal

of the American Society for Information Science 42(7)

501-514.

Gessa, N., 2007. An ontology-based approach to define

and manage B2B interoperability. Technical Report

UBLCS-2007-11. Department of Computer Science,

University of Bologna.

Gessa, N., Novelli, C., Busuoli, M. and F. Vitali, 2004.

Use and extension of ebXML business profiles for

Textile/Clothing firms. LNCS 3182. 186-195.

Guo, J., 2007. A Term in Search of the Infrastructure of

Electronic Markets. IFIP Volume 255, Springer, Bos-

ton. 831-840.

Guo, J., 2008. Collaborative Concept Exchange. VDM

Publishing, Germany.

Guo, J., 2006. Inter-Enterprise Business Document Ex-

change. In Proceedings of ICEC’06. ACM Press. 427-

437.

Lovejoy, J.(a) Pipeline Analysis (PA):How Does your

Business Flow? [online] http://www.techexchange

.com/thelibrary/Dama/pa.html

Lovejoy, J. (b) Textile Industry Data-Sharing Network: A

prototype for building bridges, using the Internet.

[online]http://www.techexchange.com/thelibrary/Dam

a/texnet.html.

MODA-ML. http://www.moda-ml.org

Leech, P., 1999. Who Says Manchester Says Cotton.

inTRAlinea 2 [online] www.intralinea.it.

Ostic, J., 1997. An Introduction to USITC Enterprise

Analysis. [online] http://www.techexchange.com/

thelibrary/Dama/An_Introduction.html.

Teng, G. and J. Hector, 2006. Integrating the US textile

and apparel supply chain with small companies in

South America. Supply Chain Management, 11(1) 44-55.

Vickrey, D., Biewald, L., Teyssier, M. and D. Koller,

2005. Word-Sense Disambiguation for Machine

Translation. In HLT/EMNLP’05, Human Language

Technology Conference and Conference on Empirical

Methods in Natural Language Processing). The Asso-

ciation for Computational Linguistics. 771–778.

ICE-B 2008 - International Conference on e-Business