Closed Bug 2057648 Opened 1 month ago Closed 17 days ago

Use a dedicated stopword list for the dnsNotFound Search CTA keyword extractor

Categories

(Firefox :: Security, task)

task

Tracking

()

RESOLVED FIXED
156 Branch
Tracking Status
firefox156 --- fixed

People

(Reporter: jbrown, Assigned: jbrown)

References

(Blocks 1 open bug)

Details

Attachments

(1 file)

The online dnsNotFound Search CTA (meta bug 2055374) derives its query in URLKeywordAnalyzer.sys.mjs, which currently relies on the shared CountVectorizer and its shared ENGLISH_STOP_WORDS list, a list also consumed by Smart Tab Grouping.

Give the Search CTA its own stopword list instead:

  • Construct CountVectorizer with stopwords disabled.
  • Apply our own curated stopword Set as a post-filter.

Our list will be a copied English base, minus a few content words the general list over-aggressively drops but which matter for this particular use case.

Assignee: nobody → jbrown
Status: NEW → ASSIGNED
Attachment #9615791 - Attachment description: Bug 2057648 - Use a dedicated stopword list for the search CTA keyword extractor r?rking,jaws → Bug 2057648 - Use a dedicated stopword list for the search CTA keyword extractor r?fchasen,jaws
Pushed by jbrown@mozilla.com: https://github.com/mozilla-firefox/firefox/commit/138198a59b88 https://hg.mozilla.org/integration/autoland/rev/a155f503b26b Use a dedicated stopword list for the search CTA keyword extractor r=mconley,jaws
Status: ASSIGNED → RESOLVED
Closed: 17 days ago
Resolution: --- → FIXED
Target Milestone: --- → 156 Branch
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: