EuroFIR UNT LATINFOODS Workshop on Production of Food Composition Data

1
Food nomenclature, classification and description
Jayne Ireland Anders Møller Danish Food
Information
2
FCDBs in a Tower of Babel

• (1) Food IDs
• In-house codes
• Simple numbers or including food classification
• (2) Food names
• National language(s)
• English names
• Scientific/taxonomic names (60)
• (3) Food classification systems
• In-house classification systems (80)(1 FCDB
  alphabetical list)
• International classification systems (10)
• Classification systems for primary foods,
  products, recipes
• (4) Food description
• Description in food name (95)
• Free-text field (40)
• LanguaL thesaurus (25 in 2005)
• Description in hierarchical code (15)
• Photographs (10)

3
Importance of food IDs in a relational database
4
Defining primary keys of tables in a relational
database

• Choose a column whose data values are unique (to
  uniquely identify each row)
• Choose a column whose data values are never
  changed
• Choose a column that does not contain any nulls
• Preferably, a column that is short and numeric
  (ex sequence number)
• Avoid choosing composite primary keys

5
Food names for identifying and retrieving
information in a FCDB

• Yes, but natural language has a few shortcomings
  
• Strawberry? Gooseberry?
• Eggplant? Grapefruit? Pineapple?
• Headcheese? Hamburg? Mincemeat pie?
• Sweetmeat? Sweetbread?
• Pudding? Black pudding?
• Catfish?
• may be inadequate and even misleading

6
Confusing terminology Common language but
different terms in the US and the UK
7
Example Endive and Chicory
8
Structured food nomenclature McCance
Widdowsons

• Apricots, raw 18 samples flesh and skin
• ready-to-eat 10 samples no stones
  semi-dried
• canned in juice 10 samples, 5 brands.
  Drained proportion 0.64
• canned in syrup 10 samples, 9 brands.
  Drained proportion 0.64
• Avocado, average Average of Fuerte and Hess
  varieties

• Pragmatically developed, informal faceted system
• Food name divided into fields to facilitate
  publication of printed food composition tables
• However, no standardised terminology

9
Structured food nomenclature USDA Qualifying
Terms

• Faceted glossary to allow systematic food names
• Terms simply listed within each facet, no
  hierarchical arrangement of term

10
Foods defined by classification systems

• Specific international food classification
  systems
• Food additives (CIAA, Codex Alimentarius GSFA)
• Pesticides (Codex Classification of Foods and
  Feeds, CCPR)
• Contaminants (Codex Classification for
  Contaminants and Toxins, GSC)
• Common Nomenclature, PROCOME, WTO, ...
• Global Product Classification (GS1 GPC, GSDN Food
  and Beverage Extension)
• European Food Groups (Cost Action 99/EFCOSUM)
  andEFSA Main Food Groups
• Food Composition Databases (EuroFIR
  classification)
• and many, many different national food
  classifications

11
Food Classification SystemsInternational systems
for exposure studies

• Based on commodities similar potential for
  pesticide residues or contaminants
• Codex Classification of Food and Animal Feed
  Commodities (CCPR)
• Codex General Standard for Contaminants and
  Toxins in Foods (GSC)

12
Food Classification SystemsInternational systems
for exposure studies

• Serve as allocation tools for food additives
• Codex Food Categorization System for food
  additives
• CIAA Food Categorization System

13
Food Classification SystemsInternational systems
for products in commerce

• Global Product Classification (GS1 GPC, GSDN
  Food and Beverage Extension) http//www.gs1.org/p
  roductssolutions/gdsn/gpc/

• data synchronisation between manufacturers and
  distributers

14
Food Classification SystemsNational and regional
classifications

• Advantages
• Reflect the culture, eating habits
• Coherent with legislation
• Disadvantages
• Food classes may not be applicable to all
  cultures
• Difficult to use on international basis

15
Food Classification SystemsNational and regional
classifications

• EuroFIR food classification
• Comparison of national food classification
  systems used in FCDBs of EuroFIR Compiler Network
• classification systems used in European food
  consumption or availability studies (Eurocode 2,
  EFG, EPIC, DAFNE)

16
Food Classification vs. Food Description

• A classification system tends to group or
  aggregate foods with similar characteristics It
  is a tool of the end-user of data.
• A variety of food classification systems have
  been developed for different purposes and reflect
  different legislations.
• A description system seeks to identify the food
  as precisely as possible, without the necessity
  of aggregating them. It is a tool of the data
  originator/ provider.

17
Food description gt more precise information

• INFOODS Guidelines for describing foods
• Faceted description designed to capture all
  available information
• Exchange of data between data sources and
  compilers of food composition databases
• Description in free text (advantage and
  disadvantage)
• Truswell, A.S., Bateson, D.J., Madafiglio, K.C.,
  Pennington, J.A.T., Rand, W.R. and Klensin, J.C.
  (1991). INFOODS Guidelines for describing foods
  a systematic approach to describing foods to
  facilitate international exchange of food
  composition data. J. Food Comp. Anal. 4 18-38.
  http//www.fao.org/infoods/nomenclature_en.stm

18
INFOODS Guidelinesinformation that collectors of
data should be encouraged to record
19
LanguaL thesaurussystematic food description,
controlled vocabulary
20
FDA "International Interface Standard for Food
Databases (IIS)

• IIS

basis for EuroFIR standard
21
Systematic food description

• LanguaL thesaurus
• Langua Alimentaria
• Multilingual faceted thesaurus
• Language-independent
• Created for use in numerical databases
• Møller A. and Ireland J. (2008) LanguaL 2008
  Thesaurus. http//www.langual.org

22
A brief history of LanguaL

• 1970s Project to develop a food description
  system for registering food manufacturers
• 1978 Expert committee to develop the vocabulary,
  known as the FACTORED FOOD VOCABULARY
• Chaired by Ann McCann, FDA
• Joanne Holden, USDA, nutrition
• Dr. Dagobert Soergel, UMd, lexicographer and
  thesaurus development specialist
• Dr. Robert Wiley, UMd, food technology
• 1981 E.Smith began using the vocabulary to index
  FDA Total Diet Study (234 foods)
• 1988/1995 FDA contracted to index many food
  databases, more than 35000 foods
• 1992 LanguaL training in Paris
• 1995 Meeting in Lahti, Finland at 2nd Intl. Food
  DB Conference Charter of International LanguaL
  Steering Committee
• 1999 LanguaL administration assumed by European
  colleagues (Cost Action 99-Eurofoods)

23
International LanguaL Charter
24
LanguaL website http//www.langual.org
25
Principles behind the LanguaL thesaurus

• An automated method for describing, capturing and
  retrieving data about food
• Based on the concept that
• Any food (or food product) can be systematically
  described by a combination of characteristics
• These characteristics can be categorised into
  viewpoints and coded for computer processing
• The resulting viewpoint/characteristic codes can
  be used to retrieve data about the food from
  external databases

26
Characteristics to systematically describe foods
Type of productFood sourcePhysical
descriptionPackagingetc.

• Each is a FACET a group of related index terms
  which represent the characteristics of a food.
• Individual index terms are called FACET TERMS or
  DESCRIPTORS. They are used to precisely identify
  each food characteristic

27
LanguaL website
28
LanguaL thesaurus alphabetical display
29
LanguaL reports available at http//www.langual.or
g/langual_downloads.asphttp//www.eurofir.org/eur
ofir/FoodDescriptionII.asp

• LanguaL 2008 The LanguaL thesaurus
• LanguaL 2008 Introduction to the LanguaL
  thesaurus
• LanguaL 2008 Multilingual thesaurus
  (English-Danish-French-German-Italian- Spanish)
• LanguaL 2008 - Documentation of changes from
  version 2007
• LanguaL Food Product Indexer 3.9 (June 2008)

30
How is food described?
Descriptors are chosen from the LanguaL
thesaurus
31
How is food described?
32
How is food described?
33
LanguaL Food Product Indexer
LanguaL Food Product Indexer
34
Index-4
Search for foods already indexed
Copy set of descriptors to new food
Visualise description of selected food
Food indexing coursesWageningen 2005, Paris
2007, Washington Bratislava 2008
35
Food files indexed by LanguaL as of August 1995
36
Recent LanguaL development

• 27 indexed European data sets
• Over 27,000 indexed foods
• Other indexed data sets
• EuroFIR BASIS (bioactive subst. in foods)
• GEMS/Foods (WHO/FAO Global Environmental
  Monitoring System) Codex CCPR
• INFORMALL (information on allergenic foods)
• USDA nutrient database
• Hungary (1995)
• French Polyphenols database (INRA) Total Diet
  Study (AFSSA)
• Updated LanguaL thesaurus
• New descriptors
• GS1 classification, health and nutrition claims
• National language versions (da, de, en, es, fr,
  hu, it)

37
Food classification and description in the
EuroFIR Standard

• Food entity
• Food code
• Food names, language codes
• Additional description

• Recipe entity
• Recipe identifier, Food code
• Recipe reference
• Recipe procedure
• Yield factors for water, fat, alcohol
• Reference for Yield factors

1-?

• Ingredients entity
• Recipe identifier
• Ingredient rank name
• Ingredient food code in FCDB
• Quantity, PCT/g
• Reference Retention factors

0-?
1-?

• LanguaL indexing
• Food code
• LanguaL code

Sampling
38
Quality assessment of food indexing in EuroFIR
network (2007)
Out of 20,000 foods, 1,000 needed to be reviewed
Physical state at room temperature (ice cream)
39
Quality assessment of food indexing in EuroFIR
network (2008)

• EuroFIR Food classification good!
• About 2 foods in data sets need to be reviewed
• Definitions need to be made clearer
• LanguaL description more accurate than food names
  in English !-)

40
Where does this lead?
USDA NUTRIENT DATABASE
FOOD INDUSTRY
EUROFIR FOOD INFORMATION PLATFORM
TOTAL DIET STUDIES
CONTAMINANTS