https://www.skysports.com being incorrect categorized as "business-ecommerce"

[kjozwiak]

Description

When visiting https://www.skysports.com, Brave will classify the website as business-ecommerce rather than sports or sports-soccer. Example:

[17424:775:0417/173224.272512:INFO:ads_impl.cc(418)] Site visited https://www.skysports.com/, immediateWinner is business-ecommerce and winnerOverTime is business-ecommerce, previous tab url chrome://rewards/

Here's the pageScore: pageScoreSkySports.txt

Steps to Reproduce

Prerequisite: build with ads enabled/working.

1. launch brave and enable brave://rewards (used 0.63.44 Chromium: 74.0.3729.75)
2. visit https://www.skysports.com and wait for the categorization

Actual result:

https://www.skysports.com is being categorized as business-ecommerce

Expected result:

https://www.skysports.com should be categorized as sports or sports-soccer

Reproduces how often:

Reproducible 100% of the time when going through the STR.

Brave version (brave://version info)

Brave	0.63.44 Chromium: 74.0.3729.75 (Official Build) beta(64-bit)
Revision	fdb7915642fef8cf997beac2554709d148e3c187-refs/branch-heads/3729@{#754}
OS	Mac OS X

Version/Channel Information:

• Can you reproduce this issue with the current release? No (not enabled on release)
• Can you reproduce this issue with the beta channel? Yes
• Can you reproduce this issue with the dev channel? Yes
• Can you reproduce this issue with the nightly channel? Yes

Other Additional Information:

• Does the issue resolve itself when disabling Brave Shields? N/A
• Does the issue resolve itself when disabling Brave Rewards? N/A
• Is the issue reproducible on the latest version of Chrome? N/A

Miscellaneous Information:

CCing @mandar-brave

[jsecretan]

This is probably just best for tracking, to revisit once @dimmu has a new model.

[tmancey]

Tested in 1.23.x and classified text with the top segment as sports-sports. This is now working as expected.

[stephendonner]

Verified PASSED using

Brave	1.23.36 Chromium: 89.0.4389.86 (Official Build) nightly (x86_64)
Revision	0c2dac31bc21e806a791641316e99920f84ea52c-refs/branch-heads/4389@{#1472}
OS	macOS Version 11.2.3 (Build 20D91)

Steps:

1. launched with --args --enable-logging=stderr --vmodule="*/bat-native-ledger/*"=6,"*/brave_rewards/*"=6,"*/bat-native-ads/*"=6,"*/bat-native-confirmations/*"=6,"*/brave_ads/*"=6,"*/brave_user_model/*"=6 --brave-ads-staging --brave-ads-debug --rewards=staging=true
2. visited https://www.skysports.com/
3. examined console logs and saw [1909:775:0312/130140.121931:VERBOSE1:text_classification_processor.cc(62)] Classified text with the top segment as sports-sports

From client.json:

{"segment":"sports-sports","pageScore":0.06054215743315853}