A chart can be translated perfectly and still become numerically misleading. Axis titles, legends and tooltips may be natural target-language prose while decimal separators, percent signs, currency symbols, compact-number abbreviations, date labels or negative-number conventions quietly change how users read the data.
Searches for data visualization localization, chart localization, dashboard localization, translate charts, localize numbers and currency, multilingual dashboards, internationalization data visualization and localized chart labels point to a specialist boundary between language and quantitative representation. Unicode CLDR provides locale data for number symbols, currencies, percentages and compact formats, while accessibility guidance requires that important data relationships not depend on graphics alone.
This guide explains how to localize dashboards, reports, analytics interfaces and charts while keeping the dataset, scale and statistical claim intact. It covers number formatting, decimal and grouping separators, percentages, currency, negative values, compact notation, dates, units, axes, legends, tooltips, label expansion, right-to-left issues, data tables, captions, accessible descriptions, colour-independent meaning and regression testing.
This article is part of eduKateSG’s Master Art of Translation architecture. It extends the professional localization layer without taking ownership away from the established translation, terminology, file-preparation, release or quality owners.
Quick answer
Keep data values and statistical transformations independent from localization. Localize presentation—titles, labels, number formats, currency symbols, dates, units and explanatory text—through locale-aware formatting and approved terminology. Never manually edit plotted values to make a chart look local. Then test both the visual and a non-visual representation of the same data.
- Freeze the data: define the canonical numeric values and transformations.
- Format: use locale-aware number, currency, percent and date formatting.
- Label: translate titles, axes, legends and tooltips precisely.
- Preserve scale: do not change units or ranges without explicit conversion logic.
- Fit: account for longer labels without truncating meaning.
- Expose data: provide accessible text/table equivalents where appropriate.
- Test: compare chart, tooltip, table and source dataset across locales.
1. Separate data from display formatting
The dataset should remain numeric or typed data rather than localized strings. Once commas, percent signs or currency symbols are baked into data, calculations and cross-locale reuse become fragile.
Professional method. Store canonical numeric values and format them at the presentation layer. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A CSV stores ‘1,234.56’ as text and is reinterpreted under a comma-decimal locale. The same value can render as 1,234.56 or 1.234,56 while remaining the identical number.
Verification. Compare the raw data object before and after locale switching. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
2. Use locale decimal and grouping symbols
Decimal and thousands separators vary. CLDR number-symbol data exists precisely so software does not assume period-decimal and comma-grouping globally.
Professional method. Use locale libraries based on CLDR/ICU or equivalent maintained data. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A translator manually swaps punctuation in chart labels. 12,345.67 may render with different separators in another locale.
Verification. Test representative small, large and fractional values. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
3. Localize percentages without changing the rate
Percent display combines a numeric transformation and locale-specific symbol placement. A raw fraction such as 0.12 may be formatted as 12%, but manual translation can accidentally multiply or divide twice.
Professional method. Define whether the stored value is ratio or already-percent and use one formatter consistently. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A dashboard turns 0.12 into 0.12% in one locale and 12% in another. The percent sign may also move or use locale spacing.
Verification. Back-calculate the displayed percentage to the stored value. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
4. Format currency from amount plus currency code
A currency value is more than a symbol. The same $ symbol can be ambiguous across currencies, and symbol position varies by locale.
Professional method. Store amount and ISO currency code separately and format under target locale and product policy. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A translator replaces ‘$’ with a local currency symbol even though the data is still USD. CLDR supports localized and international currency representations.
Verification. Check the displayed currency code/symbol against the dataset currency field. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
5. Preserve negative-number meaning
Locales can display negatives with minus signs or accounting parentheses. Changing the visual convention must not change sign.
Professional method. Use locale-aware number or accounting formats, not string substitution. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A chart annotation loses the negative sign because the style expects parentheses. A financial dashboard may show (1,234.00) as a negative accounting value.
Verification. Compare sign in source data and all displayed forms. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
6. Handle compact numbers carefully
Abbreviations such as 1K, 1M or language-specific equivalents are locale-dependent. CLDR compact-number patterns vary and some languages do not abbreviate at the same thresholds.
Professional method. Use locale-aware compact formatting and retain full values in tooltips or accessible data where useful. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A translator invents an abbreviation that readers do not use. A target locale may use a word or character for thousands rather than K.
Verification. Test threshold values around 999, 1,000, 999,999 and 1,000,000. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
7. Localize scientific notation only at the presentation layer
Scientific values have stable mathematical meaning but locale presentation can vary. Decimal symbols and exponent notation may be rendered differently.
Professional method. Keep numeric data canonical and use scientific-format support in locale libraries. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A translator edits exponent text manually and changes magnitude. 1.2×10³ and 1.2E3 represent the same value but serve different human/computing contexts.
Verification. Parse the displayed value or compare against the source number. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
8. Treat axis scale as data logic
Translation should not casually change minimum, maximum, logarithmic scale or baseline. Those choices change the visual argument.
Professional method. Keep scale definitions with chart configuration and localize only labels unless a deliberate unit conversion requires recalculation. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A localized chart changes a Celsius axis to Fahrenheit labels without converting the plotted values. Unit conversion must transform both values and axis ticks consistently.
Verification. Check plotted coordinates against transformed data. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
9. Translate axis titles with units together
An axis label often combines a quantity and unit. If the unit is lost or mistranslated, readers cannot interpret values.
Professional method. Use approved subject terminology and locale-aware unit formatting. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. The target axis says ‘Temperature’ but drops °C while another chart uses °F. A translated label can reorder quantity and unit naturally.
Verification. Trace every axis to its dataset field and unit metadata. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
10. Keep legend categories conceptually stable
Legend labels identify data series or classes. Synonym variation can make the same category appear to change across charts.
Professional method. Link legend labels to shared concept IDs and terminology resources. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. One dashboard translates ‘revenue’ three different ways in related panels. The category colour can remain stable while target-language wording is consistent.
Verification. Compare repeated categories across the dashboard. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
11. Review colour meaning independently of language
Colour may encode categories, direction or severity. Translation alone cannot fix a visualization that relies solely on colour.
Professional method. Pair colour with labels, patterns, markers or accessible descriptions where needed. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A red/green chart communicates gain/loss only through hue. Text labels or symbols can provide the same distinction.
Verification. Interpret the chart in grayscale or with colour disabled. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
12. Localize tooltip sentences as real language
Tooltips often contain dynamic numbers, dates and category names. They are small complete messages, not concatenated fragments.
Professional method. Use structured message localization as described in dynamic variables and MessageFormat. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. The tooltip is assembled as date + colon + value + unit in fixed English order. A target language may place the category or unit elsewhere.
Verification. Render tooltips with several series and long locale values. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
13. Format dates on time-series axes consistently
Time-series charts compress date labels and tooltips differently. A localized date pattern can be ambiguous when shortened excessively.
Professional method. Use locale-aware short patterns on axes and fuller unambiguous dates in tooltips or accessible tables. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. 03/04 is interpreted differently by different audiences. Month names may be clearer than numeric dates in some international contexts.
Verification. Compare axis and tooltip dates against the underlying timestamp. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
14. Allow translated labels to expand
Chart layouts often assume English label length. Long category names can overlap, rotate excessively or disappear.
Professional method. Use responsive margins, wrapping, abbreviation policies and alternate layouts rather than deleting semantic content. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. The translator shortens labels into opaque acronyms solely to fit. A horizontal bar chart can accommodate long category labels better than a narrow vertical-axis layout.
Verification. Test longest real target labels at supported viewport widths. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
15. Preserve sort order intentionally
Charts can sort by data value, time, source order or alphabetical label. Alphabetical sorting can change under locale collation.
Professional method. Record the intended sort semantics and use locale-aware collation only when alphabetical order is genuinely the goal. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A translated dashboard reorders categories alphabetically even though the source chart was ordered by magnitude. Top-ten charts should normally remain value-sorted regardless of language.
Verification. Compare the sort rule, not merely the final positions. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
16. Provide accessible table or textual equivalents where appropriate
Charts can exclude users who cannot perceive the graphic relationships. W3C WAI guidance emphasizes structured data tables, captions and descriptions to expose data relationships programmatically.
Professional method. Provide a data table, concise summary or accessible SVG/figure description according to the product and chart complexity. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. The alt text says only ‘bar chart’ with no useful information. A table can expose exact values while the chart communicates trends visually.
Verification. Navigate the equivalent using a screen reader or accessibility tree. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
17. Localize captions and summaries without duplicating noise
Accessible descriptions need to explain the chart’s purpose and structure. W3C table guidance distinguishes captions that identify content from summaries that explain organization.
Professional method. Translate the concise chart title/caption and provide additional structural or trend description only where it helps. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. The target repeats the same title three times through visible and hidden labels. A chart caption can name ‘Monthly revenue’, while a description explains that columns represent months and values are in EUR.
Verification. Review the accessibility output for clarity and duplication. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
18. Regression-test data and language together
Dashboard changes often affect both data schema and localization. A renamed series, new unit or changed aggregation can invalidate existing translations.
Professional method. Add numeric invariants, labels, locale formatting and accessible-output checks to the localization regression suite. The decision should be documented so that another translator, reviewer, designer or engineer can apply the same rule to the next occurrence without reconstructing the whole problem from memory.
Failure mode. A data field changes from ratio to percent but the formatter remains unchanged. A test can assert that 0.25 displays as the intended percentage and that the table shows the same value.
Verification. Compare chart, tooltip and table representations from one test dataset. If the result still depends on guesswork, return to authoritative locale data, source meaning, product context or the system that owns the underlying fact.
A repeatable operating sequence
A safe dashboard-localization workflow begins by freezing the meaning of the data, then localizes only the representation layers that should vary.
- Document the dataset, units, transformations and chart scale.
- Keep raw numbers typed rather than preformatted strings.
- Apply locale-aware number, currency, percentage and date formatting.
- Translate titles, axes, legends and tooltips with controlled terminology.
- Convert units only through explicit data logic.
- Test compact notation and negative-value conventions.
- Review long labels, responsive layouts and RTL where relevant.
- Provide accessible table/description equivalents.
- Compare visual, tooltip and tabular values.
- Add regression cases for every supported locale and critical dashboard.
Treat the sequence as a loop rather than a one-way checklist. If final testing reveals a problem, trace it back to the earliest decision that created it—source wording, data structure, context, template, locale rule or product logic—and repair that source where possible.
Worked scenarios
1. Revenue chart changes $ to € in translation
The translator assumes the target locale implies local currency. The controlling risk is display localization changing the economic data.
Keep the underlying USD currency code and localize only its presentation unless the product explicitly performs a currency conversion. Then test the decision under the real conditions in which the user will encounter it. Professional localization is strongest when the linguistic choice and the system behaviour confirm each other.
2. Percent axis is off by factor 100
The data source changes from percent values to ratios. The controlling risk is formatter semantics drifting from data semantics.
Document the stored unit and add a regression assertion tying one raw value to its expected displayed percentage. Then test the decision under the real conditions in which the user will encounter it. Professional localization is strongest when the linguistic choice and the system behaviour confirm each other.
3. Long translated categories overlap
A bar chart designed for short English labels receives longer German or Finnish text. The controlling risk is visual pressure causing meaning-destroying abbreviation.
Adapt chart margins, wrapping or orientation before shortening approved category names. Then test the decision under the real conditions in which the user will encounter it. Professional localization is strongest when the linguistic choice and the system behaviour confirm each other.
4. Locale alphabetical order changes categories
A dashboard originally sorted by source-language labels. The controlling risk is users assuming position reflects value when it actually reflects collation.
Decide the intended sort rule explicitly and use locale-aware alphabetical order only where alphabetical navigation is meaningful. Then test the decision under the real conditions in which the user will encounter it. Professional localization is strongest when the linguistic choice and the system behaviour confirm each other.
5. Accessible table disagrees with chart tooltip
The chart uses one formatter while the table uses another. The controlling risk is two representations of the same data communicating different values.
Generate both from the same typed dataset and locale-formatting rules, then regression-test equality. Then test the decision under the real conditions in which the user will encounter it. Professional localization is strongest when the linguistic choice and the system behaviour confirm each other.
6. Time-series axis uses ambiguous numeric dates
The dashboard is shared across multiple regional locales. The controlling risk is users reversing month and day.
Use locale-aware patterns and fuller tooltip dates, and test actual dates from the data. Then test the decision under the real conditions in which the user will encounter it. Professional localization is strongest when the linguistic choice and the system behaviour confirm each other.
Charts and data-visualization localization: twenty practice cases
These cases are deliberately varied. For each one, identify what must remain invariant, what may be localized, what evidence is required and what final test proves the result is safe to release.
1. A short label changes meaning when moved to another screen
Review the string in its actual component and user journey before changing the translation. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
2. A value is technically valid but visually unfamiliar to local users
Preserve the underlying fact and apply locale-aware display conventions at the presentation layer. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
3. A source phrase is reused for two different actions
Split the source message or add context rather than forcing one target phrase to serve incompatible jobs. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
4. A reviewer proposes a shorter translation because the UI clips
Check whether the shorter wording remains complete; if not, fix the layout instead of deleting meaning. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
5. A dynamic variable contains longer text than the demo value
Test long, accented and non-Latin values before sign-off. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
6. A user-facing number is formatted according to the source locale
Format the value under the target locale while preserving the numeric value. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
7. An automated translation is fluent but changes one proper noun
Back-check identities, names and identifiers separately from general fluency. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
8. A locale-specific rule conflicts with an official name or security term
Let the authoritative identity or security requirement win, then adapt surrounding grammar. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
9. A fallback string appears in English during one rare state
Determine whether fallback is allowed, and add the state to regression coverage. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
10. A tooltip uses different terminology from the page title
Trace both to the same concept owner and unify terminology unless the contexts genuinely differ. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
11. A source string contains an abbreviation unknown in the target market
Verify whether there is an official local form, expansion or protected global abbreviation. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
12. A translation passes linguistic review but fails after rendering
Treat rendering as part of localization quality, not as an unrelated cosmetic issue. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
13. A translated element is correct but inaccessible to screen readers
Localize the accessible name or description as part of the same communication object. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
14. A number is rounded differently in two widgets
Decide whether the widgets use different business rules or whether formatting has drifted. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
15. A new locale inherits a parent translation that is understandable but awkward
Decide whether inheritance is acceptable or whether a locale-specific override is justified. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
16. A critical message exists in three channels with slightly different wording
Preserve the same underlying event and action while adapting appropriately to each channel. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
17. A proper noun has several common target-language forms
Use the project’s authoritative naming source and document the chosen form. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
18. A visual component uses colour as the only way to distinguish states
Add textual or programmatic cues so localization does not depend on colour interpretation alone. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
19. A user-generated value contains characters outside the source team’s test set
Preserve Unicode safely and test the actual rendering/input pipeline. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
20. A last-minute source edit arrives after translation approval
Bind the approved translation to source version and reopen review if the semantic source changes. State the reason for the decision and what would make you change it. That makes the rule transferable rather than personal.
Afterwards, test the same principle against another locale, screen size, channel, identity state or dataset. A robust localization decision should survive changed conditions.
Release checklist
- Raw data remains independent from localized display strings.
- Decimal and grouping symbols are locale-aware.
- Percent scaling is defined and tested.
- Currency code and amount are separated.
- Negative and accounting formats are correct.
- Compact-number notation follows locale data.
- Axes preserve scale, unit and baseline.
- Legend categories use controlled concepts.
- Tooltips are complete localizable messages.
- Date labels are unambiguous for the locale.
- Long labels do not force meaning loss.
- Charts have suitable accessible data/text equivalents.
Frequently asked questions
Should chart numbers be translated?
The numeric value is not translated, but its human-readable formatting—decimal separator, grouping, percent sign, currency and compact notation—can be locale-specific. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
Can currency symbols change with locale?
The display symbol can vary, but never change the underlying currency itself unless the product explicitly converts the monetary value. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
What about axis units?
Translate unit labels and convert only through explicit, verified numeric logic. A label-only conversion would misrepresent the chart. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
Should charts be alphabetically resorted after translation?
Only if alphabetical order is the intended data rule. If the chart is sorted by value, time or source order, preserve that semantics. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
How do compact numbers vary?
CLDR provides locale-specific patterns because abbreviations like K/M are not universal and some languages use different words or symbols. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
What is a good accessible alternative to a chart?
Depending on the chart, a structured data table, meaningful caption, summary or accessible graphic description can expose the same essential relationships. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
How should tooltips be localized?
As complete dynamic messages with locale-aware values, not concatenated fragments. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
What should regression tests compare?
Raw values, formatted labels, units, tooltip text, table equivalents, sorting and critical accessibility descriptions. The important distinction is to separate the language decision from the underlying fact, data model or security requirement so that one layer does not silently rewrite another.
Selected references and next routes
- Unicode CLDR: Number and Currency Patterns
- Unicode CLDR: Number Symbols
- W3C WAI: Tables Tutorial
- W3C WAI: Table Captions and Summaries
- W3C WAI: SVG and image accessibility tips
Conclusion
Data visualization localization is a promise that the picture will change language without changing evidence. That requires strict separation between numeric truth and localized presentation.
When values, scales and transformations remain stable while labels, formats and accessible explanations adapt to the locale, multilingual dashboards become easier to read without becoming easier to misread.
