how could Tmdb help TM?

what is the goal of Tmdb?

how terminological pair are
added?
what is (Tmdb) called?
Introduction how cat tools show us the Tmdb matches?
The terminology database (Tmdb), also called termbase in some CAT tools, is one you build yourself
within the CAT tool. A good Tmdb optimises, refines, and customises TM recall. When the TM does not
give anticipated matches, the Tmdb can assist. Terminological pairs must be added manually. They are
not stored automatically like TUs in the TM. We create or open a Tmdb at the start of a project and add
new entries by highlighting the source and target term pairs or phrases. We access the ‘add term’
function on the menu with a right click, a shortcut code, or through an icon on the ribbon. To add
metadata, such as parts of speech (POS), synonyms or definitions, we use a different shortcut code or
click on full entry in the drop-down box. When the term recurs, matches appear in the Translation
Results window. They are marked with different icons or colours to show that they are Tmdb results. The
Translation Results window in Figure 4.1 (top left) shows from top to bottom: two red TM matches with
red ticks (1–2), three blue Tmdb matches with green ticks (3–5), and four EuroTermBank results with
their typical icon (6–9) (4.6).
Tmdb is a collection of specific words
and expressions used in a particular
subject area

(TM) is a database that stores

previously translated content in the
source and target language

which one among the three Tmdb, TM, MT should be given the priority? and why?
who added data to Tmdb? which one is more reliable TM or Tmdb?
what is the difference between Tmdb and TM?

Translators give a higher reliability rating to matches proposed by the Tmdb than TM or MT. Possible
reasons could be the specificity and accuracy of Tmdb matches, because they are not part of an auto-
assembled string (see subsegment matching in 2.2), or because they are part of added reference
material supplied by client or LSP and compensate for what is not yet stored in the TM. When translators
look for a suitable target term, the TM only recalls what has been entered and the Tmdb presents what
has been added. Therefore, if they do not know the meaning or translation of a term, they may consult
digital dictionaries, use search engines on the internet, or post queries on language or subject-specific
translator network websites.
 specificity between TM and Tmdb?

Terms and terminology how does Tmdb benefit translator?

The TM is used to memorise content and assist the translation process with the required level of
consistency, the Tmdb is a more specific guide to help translators deal with domain-related terminology
in a translation project. If the translator has access to clearly defined translations of terms in the CAT
Tmdb, it will improve quality, maintain consistency, and increase the CAT tool’s efficiency, partly through
fine-tuning the TM.

what is glossary with what is vocabulary?

Glossary example?

A glossary can be defined as a language dictionary in one or more languages containing all the
terminology of a domain usually preferred by an organization. A vocabulary is a list of terms with
synonyms and definitions (or explanations) in two or more languages, relating to a specific subject field,
but not pertaining to the activities of a company or organization like the glossary. A good example of a
glossary created by an organization is the Online Glossary on Governance and Public Administration
developed by the Division for Public Institutions and Digital Government (DPIDG), of the UN Department
of Economic and Social Affairs in collaboration with CEPA**(link: glossary), Classification of
Environmental Protection Activities: the purpose of glassory?

The purpose of the Glossary is to provide United Nations Member States, and all other interested
parties, with a common definition of the basic terms and concepts related to governance and public
administration used in United Nations documents, and DPIDG. In particular, the Glossary aims at
improving the clarity of the intergovernmental deliberations of the United Nations itself; and at assisting
Member States to better implement United Nations resolutions by providing a more unified
understanding of governance and public administration terminology.
why UN glossary is designed?
The UN glossary is monolingual and contains definitions only, specifically designed to set terminology
standards which will be recognised by all member states. The UN glossary is monolingual and contains
definitions only, specifically designed to set terminology standards which will be recognised by all
member states. T

what is the terminology of a subject?

what is term extraction? what is term and concept?

Term extraction

Term extraction is the process of extracting domain-relevant terms from a corpus, a body of texts or
words. Before we extract terms, we should at least know how to identify and categorise them. A term is
a word (simple term) or a multiword expression (complex term) of a concept within a specialist subject
(domain). A concept is an idea that corresponds to its essential features. It will cover many varieties,
such as pencil, piece of chalk, pen, all products of the same concept, the writing tool. The terminology of
a subject is the set of special words and expressions used in connection with it (Cobuild). Terms behave
in a distinct manner and have relatively fixed contextual surroundings (4.4). In the English language these
are typically noun phrases, such as ‘adapter socket wrench manufacturer directory’, or verb phrases,
such as ‘to be in/out’, which in cricket terminology refers to whether the players are batting or not. They
are difficult to translate. Other features may be typographical, e.g. italics for scientific names, an
 acronym, such as WYSIWYG (What You See Is What You Get) or initialisms, for example, UN (United
Nations, but known as UN), or the official title of a position (CEO – chief executive officer), organisation
(FC – football club), department (DS – department of state) or unit (ARMAD – Armoured and
Mechanised Unit Air Defence). how different CAT offer term extraction?
where are extracted terms saved?
In a CAT tool, we can extract term pairs by highlighting and adding them to Tmdb on the fly. Several CAT
tools**(link: Term Extraction) offer monolingual term extraction as a regular part of the translation
process or as a separate feature to build up glossaries or Tmdbs (Zetzsche 2003–2017). Term extract is a
very useful but often underused feature. It extracts important words and expressions in a new source
text or a corpus of existing translations. The selection is data-driven and uses algorithms. Consequently,
a term list may contain a quite a few false positives, which are irrelevant results due to multiple
meanings. An extracted term list, however, enables a translator to research important terms before
translating the text and to include them in a Tmdb if they appear to be correct and consistent.
how CAT tool construct term list?
The CAT tool with an external Tmdb (SDL Trados Multiterm) uses existing translated documents saved in
the CAT tool and checks the frequency of terms at subsegment level to create a term list. In CAT
programs with an integrated term extraction editor, we can run term extraction each time we add new
files. It will simultaneously search our Tmdbs. It is advisable to edit and maintain the lists constructed by
the CAT, because they are created as proposals and may need some reconfiguration. Filtering (2.5.1) is
also recommended to prevent the extraction of syntactic words without terminological significance, such
as articles (the, a/n). Before we apply the useful function of term extraction, we need to understand how
to create a reliable and usable Tmdb. The CAT tool can do the job, but it works better if we add
metadata. They operate as a filter and reduce the number of useless matches.
where does metadata appear?
what is the difference between metadata TM and Tmdb?
Metadata where is metadata saved? If the Tmdb is well filled and edited is
called?
If we compare the translation memory with the terminology database, there is a significant difference in
the amount and type of metadata either can store, and our control over them. Metadata in the TM tell
us when and by whom a segment was translated or edited. This can be useful when comparing
previously translated versions, which are all stored in the TM. When we set up the project, we can state
the domain, client, and deadline. The Tmdb, in contrast, is filed manually and metadata are added by the
translator: SL/TL variant(s), fields of application, part of speech (POS), definitions, sources, usage, and
context (2.2, 4.4). When a Tmdb match appears in the Translation Results window, we can open a dialog
with a right click to view added metadata. If the Tmdb is well filled and edited we call it a ‘high-end
termbase’ (Melby 2012). This means that the concept entries have been checked (not taken from the
web without quality control), and are accompanied by equally checked information, the metadata. This
work practice has a positive impact on the quality of the TM and reduces TM edit time if appropriate
terms are selected immediately and consistently. Table 4.1 shows how a comprehensive terminological
entry in the database removes ambiguities and supports terminological consistency in a translation
project. It gives an example of a definition of a special breed of cat in the Tmdb of a 20,000-word-
encyclopaedia-on-cats translation project.
how could comprehensive terminological entry
effect translation?
In Figure 4.2 above (memoQ 9.2) The translator completes the following:

• entry – source term(s): [+] for additional terms, [|] for change, and [-] for

removal

• matching:

• matching ‘50% prefix’: morphological changes (4.5) such as verbal

prefixes or endings are not immediately identified as matches by the

TM. The ‘50% prefix’ is the default setting. In this case, the Tmdb will
look for ‘supplies’ and ‘supplied’ but not for ‘supplying’. The default can

be changed to fuzzy, exact, or custom

• case sensitivity: this can be set to ‘yes’ or ‘no’, if you would like a term to

contain small letters or caps or both.

• usage enables us to add forbidden terms, meaning how a term should not be

translated

• grammar enables us to add POS, gender, and number. For example, the term

‘research’ can be a verb or a noun

• definition (Table 4.1)

Definitions are sometimes needed to clarify the meaning and standardise the

term (4.1.4, 5.4), particularly in technical texts. A good and helpful definition will

meet the following criteria:

• Predictability – the definition inserts the concept into a concept system: a

carrot is a vegetable

• Simplicity – the definition is clear and concise and no longer than one sentence

• Affirmativeness – the definition states what the concept is

• Non-circularity – the definition does not begin with the term and does not

use words which are dependent on the concept

• Absence of tautology – the definition is not a paraphrase of the term, but gives

a description of the semantic features

• Part of speech – the definition begins with the same part of speech as the term

you are describing

why should we apply standardization?
what is standardization ?

Standardisation and quality assurance why Terminology standards is important?

It is best to follow agreed terminology standards and to perform regular terminology maintenance. The
industry relies on technical specifications and criteria, rules and guidelines to ensure that materials and
products are interconnected and interoperable. This is called standardisation. We need to apply similar
standards to terminology so that we agree which technical terms will be accepted as standard.
Terminology standards for preferred terms and definitions ensure a common understanding of the key
concepts.
ISO standards**(link: ISO terminology) can be applied to multilingual or bilingual terminology databases
(5.4) and termbanks (4.6). But how are ISO standards of terms determined? Terms need to pass a
reliability assessment; they should be able to cope with emerging knowledge and not be obsolete on the
day they are standardised. Acronyms and names of school examinations, for example, are often
reviewed, modified, or replaced. Our CAT terminology databases need to be maintained and kept up to
date accordingly.

There are terminology management tools available to help us check consistency and relevance. They
also provide quality assurance and a certain level of terminology management **(link: terminology
management). We can import bilingual CAT formats, such as TMX, XLIFF in the terminology management
tool that will then run checks for completeness, consistency, number, and tags. Results are presented in
a common results window. The tool can also perform Quality (QA) checks in the CAT tool.
the insertion of target terms in the CAT editor
what is the difference between jonuir and senuoir translator? besed on what?
which approach is better?
Web searches
When adding target term retrieved from the internet what should check?
In their terminology searches, translators make prolific use of online mono/bilingual dictionaries,
specialised glossaries, comparable and parallel corpora (4.3), MT, social networking and cloud
technology (Wang and Lim 2017). Wang and Lim (2017) found that senior translators rely more on
experience and memory and have a more linear and syntagmatic approach to reading the ST. The senior
translators read from beginning to end, compared to junior translators who are more paradigmatic and
focus on the segment and its alternatives. The paradigmatic approach is more likely to lead to
misinterpretations, When we add a target term retrieved from the internet to the Tmdb, we must be
sure of its validity and accuracy. Furthermore, the insertion of target terms in the CAT editor needs to be
based on our understanding of the entire source text, and our translation benefits from a linear,
syntagmatic approach.

is The World Wide Web (WWW) reliable for term searchs?

Search techniques what types of engine should we use to search for terminology?

The World Wide Web (WWW) is open source, it can be accessed free of charge, it can be modified,
censored, and manipulated by users. The internet was not designed for terminology searches. Hence our
online term searches must be sophisticated. Web searches for terminology can be enhanced if we
choose appropriate search engines best suited for our term searches. There are many search engines,
but the main categories are threefold: the main categories of search engines?

• (General-purpose) search engines are software applications that help search for

websites in WWW and then access and display them. Each search engine has

its own special features which also vary per country

• Specialised search engines show index pages for special topics only, which may not

be shown in general-purpose search engines

• Meta-search engines are search tools that operate as aggregators and use other

search engines’ data to produce their own results**(link: meta-SE)

 What are search engine operators? and What is the difference between
them?

Our search techniques can be made more term specific by using the operators

AND, ANDNOT in the search box of the search engine. These operators are called

Boolean logic. Its use has distinct advantages and immediate results:

• AND narrows the search by retrieving only documents that contain the

keywords we enter, e.g. ‘tea AND coffee’ excludes ‘tea and chocolate’,

• OR expands the search by returning findings with either or both keywords, for

example, ‘nursery OR day care centre’, and NOT limits the search to the first

keyword only, for example ‘coffee ANDNOT tea’. Boolean logic

does The quality of the website correlate the quality of its
what is A domain suffix? terminology?
how URL help us? com, org, co.uk are domain suffixes entirely reliable?
Boolean logic is a good filter and makes hits more specific. Another filtering method is to check the
source and only accept term hits from reliable URL website addresses. The URLs presented to us by the
search engine tell us about the origin of the website and help us determine a level of reliability.
A domain suffix**(link: domain suffix) is the last part of a domain and defines the type of website we
may be about to access. ‘Com’ domains are commercial websites, whereas ‘org’ domains are used by
organisations, and ‘co.uk’ means that the business is run in and from the UK. Domain suffixes are not
entirely reliable and the actual website must be checked to be sure that it was not hijacked by a less
trustworthy body for their own purposes. The quality of the website does not necessarily guarantee the
quality of its terminology and linguistic checks are necessary

what is corpora? should digital corpora be integrated in CAT tool?

Corpora what constitute corpora in CAT tool?

‘Corpora’ (plural form of ‘corpus’) are large structured sets of texts or words. A terminology database in a
CAT tool is a corpus of words, like a dictionary; the TM is a corpus of phrases or sentences. Aligned texts
and reference files constitute corpora in a CAT tool (Alignment 2.3.1 and Reference files 2.3.2). Digital
corpora can be external or integrated in the CAT tool.

difference between digital dictionary and reputable dictionary?

Digital dictionaries who determines the quality of digital dictionary?

Dictionaries are corpora. A reputable dictionary will ‘sample’ a term before it is entered, which means
that the lexicographers will look at one or more screenfuls of relative citations to see how the term
behaves grammatically and lexically before it is entered or after an existing entry has been revised. With
the vast rise of digital dictionaries, unpublished and free of charge, we cannot be sure that similar
sampling procedures are applied, and consequently the quality of many digital dictionaries must be
determined by the user.
what are digital dictionary advantages?
Digital dictionaries offer new and excellent opportunities for lexicographers to compile corpus-based
dictionaries (Müller-Spitzer and Koplenig 2014). They can be updated online without printing costs, the
entries can be linked to collections of texts (4.4) and multimedia, such as audio files, graphs, and images.
 what is Wikipedia?
what does reliability mean?
what are the criteria of digital dictionaries?
In addition, the entire internet community can contribute, with Wikipedia as an example. Wikipedia has
become a popular website, used as an immediate source of information. However, its open-source
character, meaning that anybody can add, remove, and modify content without any validation or control,
makes its content unstable, if not unreliable. The study by Müller-Spitzer and Koplenig (2014) showed
that the following criteria were considered important in digital dictionaries by their
respondents: reliability of content, up to date, and to a lesser degree, accessibility and clarity. Reliability
of content means that all details represent actual language usage and that all details are sampled and
validated on a corpus how can we determine the source and origin of
digital dictionary? can digital dictionary be added to CAT tool?
Digital dictionaries are like websites: we can determine their source and origin by checking their URL,
but without an ISO standard, they do not have a quality label. If bilingual dictionaries are machine
translated, they are even less likely to have been validated. Digital dictionaries are generally protected
and it is not possible to download and import them in the CAT tool. They must be treated as
independent TEnTs.

is Multilingual data corpora TEnTs?

Multilingual data corpora does multilingual data corpora show us if the transaltion wrong or not reviwed?

Many large organizations make their multilingual corpora available online for general use and sharing,
which means that users can upload and download terminology to and from the websites. There are
some very impressive data sources available on the web, which we can use as TEnTs. The programs use
web crawlers to search for translated content online, including e.g. EU multilingual data and then match
it with web-based dictionaries. In Figure 4.3 we see the results if we enter a source term or an entire
phrase in the tool: it displays a vast amount of complete target segment matches. Some matches have
warning symbols, such as ‘this translation can be wrong’. The website of the source is listed below the
match. The tool also states if translation pairs have not yet been reviewed.
 is TAUS multilingual corpora reliable? What is TAUS Multilingual corpora?
Multilingual corpora with a more reliable quality are provided by TAUS. Its Data cloud**(link: multilingual
corpora) is an extensive collection of translation memory data from large translation buyers
predominantly in software and IT industries (Zetzsche 2003–17). It is a neutral and secure repository
platform and gives translators, language service providers, developers, researchers, and stakeholders
access to a number of services, such as Search, Upload (your own data for sharing), Discover &
Download (only if you have uploaded, and acquired credits), Account History, My Data, and Feedback. Its
collaborative plan gives you credit points if you upload data and you gain credits in return which enable
you to download. The Data cloud is accessed through the web or API, and accessibility depends on a
membership plan. Once corpora are created, they are offered as ‘libraries’ to other potential purchasers.
The i (information) icon after the first entry in Figure 4.4 gives information on industry (Legal Services),
type (Standards, Statutes, and Regulations), Data owner (European Parliament), Data Provider (TAUS).
Google offers us multilingual data corpora by letting us search the web in the requested language. It
gives a random set of links to the search term in context. There are many other free downloadable
parallel corpora of texts (for alignment) and terms available to fill your tools, but this does not guarantee
the quality of your translations (Zetzsche 2003–2017). Their web mining for material is done
automatically and manual corrections are not necessarily carried out
How should we use free online multiligual corpora?
The quality of online multilingual data does not necessarily meet guaranteed standards or required
criteria. It does not mean that we cannot use them, but it implies that we must have a system in place to
check their usage and accuracy.

The creation of corpora how to add new data in corpus?

An established dictionary published online may have a history of 200 years. It can then draw on a wealth
of reliable, tested, and informative material. Its lexicographers will use language databases, i.e. corpora,
in multiple languages to extract terms and try and test usage. The corpora ensure that language in the
dictionary is authentic and accurate. New data is fed regularly from spoken or written material, from
television, radio, websites, newspapers, journals, and booksIf we want to create our own corpora to
build our own terminology databases, we can extract terms from reference materials by ‘bootstrapping’
our corpora onto existing and available web corpora. For this there is the webBootCaT technology and
there are BootCaT tools. The designers (Baroni and Bernardini 2004) of the webBootCaT built a tool that
does not need downloading but can be created by using the Google search engine. The basis method is
that you first select a few seed terms (keywords), then send queries with the seed terms to Google and
then collect the pages in Google’s hits. The vocabulary in your created corpus of texts can be extracted
automatically in the tool. The webserver of the designers will hold the corpus and load it into a corpus
query tool, such as Sketch Engine where it is investigated and analysed. The terminology database you
have created must be converted to suitable formats (CSV) to import them in CAT tools. The corpus query
language in the term extracting tool is a special language that looks for complex grammatical or lexical
patterns to establish suitable search criteria. We can see the outcome of its searches in the concordance,
where our queries are presented in context. does CAT concordance give multiple
translation?
is CAT concordance bililgual?
is it seen as substitute for Tmdb? does CAT concordance give definitions?
CAT concordance

It is intended for term or phrase searches and is often seen as a substitute or an alternative for the
Tmdb. It allows you to access the TM when searching for individual terms or phrases within string. The
CAT concordance retrieves all the entries of a specific word or phrase from the TM and displays them
bilingually so that context-based information can be given of source and target segments. In the CAT
concordance we must take account of subtle differences due to segmentation or morphological changes
which affect what is retrieved by the TM). The CAT concordance is not only bilingual but also parallel as it
gives multiple translations for one source term or phrase. A concordance does not give any definitions,
but it recalls nuances, and different shades of meaning in context. If the Translation Results window does
not propose any matches, the translator can retrieve term pairs through the concordance function. If the
term pair is embedded in a string that is different to a previously confirmed string, the TM does not
propose a match. The concordance function, however, ignores the string and focuses on the term Pair.
does CAT concordance show a result that in Tmdb?
The limitation of the CAT concordance is that it can only recall what has been entered in the TM. If the
Tmdb presents matches from imported reference files, which have not yet been confirmed and entered
in the TM, the concordance will not offer any results.

what is the advantages of external

External concordances what are the types of external concordance?
concordance?
There are two types of external concordance: those which enable us to create a what it can do?
and how that
concordance through reference files, and comprehensive web concordances that we happen?

may use to check our target terms. An external concordance can automate the process of finding specific
reference texts on the web and collate them in one corpus (Olohan 2004). The process is as follows: first
the user provides a list of key terms and phrases in the program, which then sends them as queries to a
search engine that returns a list of potential URLs. The user selects suitable URLs, and the program
retrieves the webpages and converts them to plain text or saves them in ‘txt’ format. Some CAT tools
accept the text files as reference materials, and they will then be incorporated in CAT concordance
searches. The external concordance can retrieve and process many texts in a short period of time which
makes it a very useful TEnT. Its external nature may have an additional advantage. Reference materials
may not need keeping. They are useful to understand the ST or find translations for the TT as a one-off.
TUs are then entered in the TM, which makes new terminology entries retrievable in the CAT
concordance without having to store the reference material. The CAT term extraction editor (4.1.2) can
add them to theTmdb.

Digital dictionaries contain billions of words. If we search high-frequency words like ‘make’, the British
National Corpus (BNC) may give hundreds of millions of results (known as ‘citations’). If translators want
 how good external concordance help translators?

to know more about actual usage, they can use online corpus search software which displays the
searched term or phrase in combination with other words, called collocates, and the software gives the
frequency of the collocations. A good external concordance can assist a translator in the following ways:
show the most typical combinations of the word (or phrase), show synonyms (thesaurus), compare with
another word and find examples of usage. In a parallel concordance we will find examples of alternative
translations. We can apply filters to the tool, so that it customises and collates our searches. Because
sources are provided for each citation, we can tell the program which sources are preferred.

Morphology and the terminology database

what should the translator do to solve the problem of morphology?

It can be frustrating when neither TM nor Tmdb recognises a term due to an added prefix or suffix, while
you are convinced you have previously translated and confirmed it in one or both databases. This is a TM
problem related to morphology, i.e. the form, shape, or structure of the word. If a derivative is not
recognised, translators can try to resolve this by using filters (2.5.1) or highlighting the core of the word
without the prefix or suffix and search in the concordance feature. TEnT manufacturers have created
different solutions to prevent TM non-recognition associated with morphological hange. A translation
management system (7.5) uses AI to discern morphological changes in words. Some CAT tools require
the user’s application of an asterisk or pipe character to tell the tool that following characters should not
exclude possible matches, such as ‘support*/|’ must include ‘supported’, ‘supporter’. A CAT program
with an external terminology database has incorporated morphology for a range of languages. The
manufacturer decided not to use the external Tmdb but moved the process into the cloud where it is
combined with the program’s MT service**(link: morphological change). The user can select an existing
webbased Tmdb and add new term pairs. Searches in the Tmdb differentiate between linguistic and
character-based fuzzy searches. The former is based on the morphological core of the term and the
latter is based on searches of three letters or more, comparable with asterisk and pipe-based searches.
Especially interesting is the manufacturer’s use of third-party tools, i.e. an existing web-based
‘morphology engine’, instead of developing their own engine. It should be noted that if the webbased
Tmdb is shared, it could impact confidentiality.

Zetzsche (2019 (268th Journal)) stresses the need for morphological support in all CAT tools. Particularly
languages such as German, with compound words would benefit from improved recognition. It would
also increase the accurate verification of terms in the quality assurance (QA) function (5.2.1). The CAT
tool may flag up false positives during a QA check and mark morphological changes as supposed
inconsistencies or false matches. Until all CAT programs support morphological changes, we have the
option to use the filters and change fuzziness settings. Although it is not advisable to lower the
percentage of fuzzy matches, it could be useful in the read-only mode during a revision as it is more
likely to include some morphological changes.
 what are termbank's approaches?

is the majority of termbank integrated?

Termbanks
what is termbank?

Large external terminology databases are often referred to as termbanks**(link: termbanks). They have
different modes of operation within or outside the CAT tool. If an integrated termbank in a CAT tool
presents termbank information, or matches, in a results window while we are translating, it is called a
‘push’ approach, whereas if we use the concordance or consult external terminological material, it is
called the ‘pull’ approach (Warburton in Chan 2015: 656). One CAT tool performs push lookups through
the EuroTermBank online terminology service, which is a databank with more than 600,000 entries of
more than 1,500,000 terms. Figure 4.6 shows how a click on the Quick Access tab displays the Term
Lookup icon in the CAT program. Because the Euro TermBank is integrated in the tool, the user can
toggle between ‘Search memoQ term bases’ and ‘Search Euro TermBank’. One unsuitable result for
‘power’ is shown in the Lookup term dialog. The Translation Results pane does not give a Tmdb result,
and the remaining option is to click on the Online Lookup in the Lookup term dialog. The subject field
was not selected, which accounts for an inappropriate result found in the subject field of
‘Communications and Education’. The TM, however, gives an acceptable 101% context match. The
different features complement each other well, depending on the smart use of the translator.

is external termbank reliable? example of Canadian termbank?

External termbanks can be accumulated and aggregated on behalf of governments, multinational

organisations, or international umbrella organisations, such as the UN, which rely on ISO standards. It is
possible for registered users to download, but not to upload or share, terms in contrast to many online
corpora run by web-based companies which make a profit through sharing and advertising. TERMIUM
Plus®, for example, is maintained by a translation bureau under the auspices of the Canadian
government. It is one of the largest termbanks in the world and it gives access to millions of terms in
English, French, Spanish, and Portuguese**(link:Termium). It complies with ISO standards on
Terminology (5.4). The majority of the termbanks are independent standalone tools not integrated in
 how to solve the problem of terminology resources
are not integrated?

CAT tools. Termbanks are valuable components for us to use as terminological resources. Several CAT
tools have the Term Extraction editor feature (4.1.2). Unfortunately, many termbanks and other
terminology resources lack integration in the translator’s workflow and the translator will need more
than one screen to open several windows simultaneously. Standalone term extraction tools remove this
problem and may suit translators who are keen to tap into as many resources as possible and build their
own customized Tmdbs. what are our quality and quantity of TM
determined by?
Today’s translator cannot complain that they are short of terminology resources. Our content and quality
control over the TM are determined by quality and quantity in our Tmdbs, more than we realize. But we
must first create and build terminology databases.