v1.48.0

Added

• Improves the feedback for the assessment that checks the length of the text in cornerstone articles.

Fixed

• Fixes a bug where a Flesch reading ease score of exactly 90 would incorrectly trigger a red bullet.

v1.47.0

Fixed

• Fixes accidental removal of the SEO Assessor export from the index. Props Kingdutch.

v1.46.0

Added

• Adds readability analysis for Swedish.
• Adds prominent words for Swedish.
• Improves keyword recognition in Swedish by filtering function words.
• Improves the transition word assessment for German.
• Improves the error logging when there is an error in an SEO or readability assessment.

v1.45.0

Fixed

• Fixes a bug where special characters from certain word lists weren't correctly escaped when matched with a regex. This resulted in eggs being incorrectly matched as the transition word e.g., for example.
• Fixes a crash when loading the createWorker module because the window was accessed in the global scope immediately.

Added

• When the recalibration feature flag is switched on:
  • The single title assessment is added. This assessment makes sure that you don't use superfluous H1s in your text.
  • The following assessments are not used anymore:
    • The assessment checking the length or your URL.
    • The assessment checking whether your URL contains stopwords.
  • Assessments changes:
    • Keyphrase density: changes scoring schema to account for the length of the keyphrase and changes feedback strings so that we give feedback about the number of occurrences rather than a percentage.
    • Outbound links assessment: changes the scoring schema so that red bullet instead of an orange bullet is shown when you have no outbound links.
    • Image alt attributes: if there are at least 5 images, checks whether the alt tags contain the keyphrase or synoynyms in 30-70% of all images. If there are less than 5 images, 1 image with the keyphrase or synonym in the alt tag is still scored as good.
    • Keyphrase in title: function words preceding the exact match keyphrase are ignored when determining the position of the keyphrase in the title.
    • Keyphrase length: makes the scoring scheme less strict for languages that don't have function word support, so that for these languages keyphrases with 1-6 words are scored as green, 7-9 as orange, and more than 9 as red.
    • Keyphrase in subheading: only takes H2 and H3 level subheadings into account and changes the scoring schema so that 30%-75% of these subheadings need to include the keyphrase or its synonyms. In languages without function word support, a match is only counted if all the words from the keyphrase/synonym appear in the subheading.
    • Text length: on taxonomy pages, the recommended minimum text length is increased from 150 to 250 words.
• The browser console now shows more descriptive error messages when something went wrong during analyses in the web worker.

Changed

• Improved README Usage to detail the Web Worker API.

v1.44.0: Release 1.44.0

Fixed

• Fixes a bug where keyphrases weren't recognized in the URL when the words in the URL were separated by underscore characters instead of hyphens.
• Fixes a bug that caused numbers to be stripped when marking a keyphrase containing a number, e.g. Yoast SEO 9.3.

Added

• Adds relevant words from the browserified example to the dev tool.
• Adds the option to use local morphology data in the dev tool.

Changed

• Improves error handling in the analysis web worker by rejecting the last request instead of just throwing an error.

v1.43.0

Fixed

• Fixes assessments failing when using a < sign in the content.
• Fixes a bug where paragraphs were sometimes not correctly detected because paragraph tags were not automatically added in WordPress-like fashion.

Added

• Adds toggles to use the different assessors in the webpack example (e.g., for all the different combinations for cornerstone and taxonomy pages and related keyphrase).
• Introduce logger in the AnalysisWebWorker to replace the development console log.

Changed

• Refactor SEO assessment file names and exports. Props Kingdutch.
• Disables the non-functioning markers for the subheading distribution assessment.

v1.42.0

Fixed

• Improves keyword recognition in the first paragraph on texts which start with images and captions.

Other

• Removes non-functioning eye-markers from the link keyphrase assessment.

v1.41.1

Fixed

• Fixes a bug where the Chrome browser tab would crash on Windows when a French or Italian text contains sentences in passive voice, props CarloCannas.
• Fixes a bug where the Yoast SEO analysis would error if used together with the DelightfulDownloads plugin.

v1.41.0

Added

• Introduces two new principles for keyword recognition:
  • Makes keyphrase recognition flexible with regards to word order. This means that the keyphrase SEO WordPress plugin will be found in the sentence This is the most popular SEO plugin for WordPress. In order to use exact matches, the keyphrase can be enclosed in quotation marks.
  • When matching keyphrases for the various assessments, the analysis only targets content words and ignores function words (e.g., the or and). This functionality is available in English, German, Dutch, French, Spanish, Italian, Portuguese, Russian and Polish.
• Implements support for word form recognition for keyphrases in English (requires Premium configuration).
• Improves the feedback texts for all SEO and readability assessments.
• Adds functionality to append a query string to the assessment links through the analysis worker.
• Adds an assessment that checks whether your keyword consists only of function words.

Changed

• The analysis of the following assessments incorporates the new keyword recognition principles:
  • Image alt attributes: checks whether there’s at least one image with an alt tag that contains words from the keyphrase. An exact match isn’t required anymore.
  • Keyphrase in introduction: checks whether words from the keyphrase are matched within one sentence in the introduction or, if not, whether they are present in the first paragraph at all. An exact match isn’t required anymore.
  • Keyphrase in title: still checks whether an exact match of the keyphrase is found in (the beginning of) the title, but now also recommends improvement if all words from the keyphrase are found in the title.
  • Keyphrase length: has new boundaries to check whether the keyphrase is not too long. For languages that have support for function word stripping (see above), only content words are taken into account.
  • Keyphrase in meta description: checks how often all words from the keyphrase are matched within the meta description.
  • Keyphrase in subheading: checks whether at least one subheading contains more than half of the words from the keyphrase. An exact match isn’t required anymore.
  • Keyphrase in slug: checks whether a sufficient number of words from the keyphrase is used in the slug. The number of words required depends on the length of the keyphrase.
  • Keyphrase density: checks whether there are enough keyphrase matches; a match is defined as a sentence that contains all words from the keyphrase. The upper boundary for a good score is higher when word form recognition is available, since in that case the analysis is able to pick up more matches.
  • Link focus keyphrase: the assessment that checks whether you’re using your keyphrase to link to a different article doesn't require an exact match anymore.
  • Keyphrase distribution uses an improved algorithm that checks whether the keyphrase is evenly distributed throughout the text.
• The following assessments will also count synonym matches as keyphrase matches:
  • Image alt attributes
  • Keyphrase in introduction
  • Keyphrase in meta description
  • Keyphrase in subheading
  • Keyphrase distribution
• Deprecates the assessment that checks if stopwords are used within the keyphrase.
• The analysis for related keyphrases only shows assessments relevant for the specific keyphrase. It omits assessments that are non-keyphrase-related and assessments that should only be applied to the focus keyphrase.

Fixed

• Fixes a bug where the keyword would not be found in the slug when containing punctuation, for example the keyphrase apples & pears in the slug apples-pears.

v1.40.0

Bugs:

• Fixes a bug that caused keywords beginning with the Turkish characters İ / i and I / ı not to be recognized when changing that character from lowercase to uppercase and vice versa.

Enhancements:

• Exposes word boundaries for use in other libraries or applications.

Changed:

• Drops TypeScript support.
• Changes all usage of lodash to lodash-es.
• index.js has been rewritten to use ES6 Module import and export logic, instead of Node's require logic.