Classify search engine results page content #6000

jsecretan · 2019-09-12T21:58:58Z

No description provided.

LaurenWags · 2022-06-29T13:47:34Z

Verified with

Brave | 1.41.82 Chromium: 103.0.5060.66 (Official Build) beta (x86_64)
-- | --
Revision | 20b1569438a85e631d15e83eb355e3e326e5da6f-refs/branch-heads/5060@{#1066}
OS | macOS Version 12.4 (Build 21F79)

Verified test plan from brave/brave-core#13522 for a selection of the sites listed (not all).

https://github.com/ - PASSED

Using 1.40.107
- Launched a clean profile and enabled ads.
- Waited for all items to complete and components to download.
- Did a search for "ruby".
- Confirmed no mention of Classified text with the top segment as... after performing this search.
- Did see the below in my logs:

[79627:259:0629/123826.460466:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

Repeated the above using 1.41.82.
- After searching github.com for "ruby", confirmed the below shows in my logs:

[77166:259:0629/101652.316679:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as technology & computing-programming

https://amazon.com/ - PASSED

Using 1.40.107
- Launched a clean profile and enabled ads.
- Waited for all items to complete and components to download.
- Did a search for "espresso machine".
- Saw mention of Classified text with the top segment as... after performing this search.

[78961:259:0629/122141.826300:VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as home-appliances

Repeated the above using 1.41.82.
- After searching amazon.com for "espresso machine", confirmed the below shows in my logs:

[79039:259:0629/122241.965933:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as home-appliances

https://www.semanticscholar.org/- PASSED

Using 1.40.107
- Launched a clean profile and enabled ads.
- Waited for all items to complete and components to download.
- Did a search for "economic growth".
- Confirmed no mention of Classified text with the top segment as... after performing this search.
- Did see the below:

[79287:259:0629/123445.689539:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

Repeated the above using 1.41.82.
- After searching semanticscholar.org for "economic growth", confirmed the below shows in my logs:

[79488:259:0629/123736.755781:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as business-business

https://www.webcrawler.com/ - PASSED

Using 1.40.107
- Launched a clean profile and enabled ads.
- Waited for all items to complete and components to download.
- Did a search for "monstera".
- Saw mention of Classified text with the top segment as... after performing this search.

[79883:259:0629/124430.419159:VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as home-garden

Repeated the above using 1.41.82.
- After searching webcrawler.com for "monstera", confirmed the below shows in my logs:

[79972:259:0629/124631.194934:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as home-garden

https://www.amazon.co.uk/ - PASSED

Using 1.40.107
- Launched a clean profile and enabled ads.
- Waited for all items to complete and components to download.
- Did a search for "anker charger".
- Saw mention of Classified text with the top segment as... after performing this search.

[85368:259:0629/155720.006724:VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as home-appliances

Repeated the above using 1.41.82.
- After searching amazon.co.uk for "anker charger", confirmed the below shows in my logs:

[85565:259:0629/160111.840785:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as cell phones-cell phones

stephendonner · 2022-06-30T21:51:43Z

Verification `PASSED` using

Brave	1.41.86 Chromium: 103.0.5060.66 (Official Build) beta (64-bit)
Revision	20b1569438a85e631d15e83eb355e3e326e5da6f-refs/branch-heads/5060@{#1066}
OS	Linux

Shared Steps:

installed 1.41.86
launched Brave
loaded the base URL noted, for each
searched for the quoted term
checked console logs for text-classification snippet

duckduckgo.com - "shoes"

1.40.107

[14081:14081:0630/144804.244637:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.86

[14463:14463:0630/144939.553472:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as fashion-fashion

search.brave.com - "barbie dolls"

1.40.107

[18922:18922:0630/170221.055122:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.86

[19797:19797:0630/170514.302088:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as other-other

stackoverflow.com - "python"

1.40.107

[20212:20212:0630/170859.943418:VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as technology & computing-technology & computing

1.41.86

[21522:21522:0630/171410.003110:VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as technology & computing-programming

twitter.com/explore - "covid"

1.40.107

[21939:21939:0630/171607.933452:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.86

[22368:22368:0630/171759.284545:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as travel-adventure travel

excite.com - "ukraine"

1.40.107

[22784:22784:0630/172129.282665:VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as history-history

1.41.86

[23205:23205:0630/172314.808215:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as history-history

search.yahoo.com - "news"

1.40.107

[23690:23690:0630/173019.199762:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.86

[24110:24110:0630/173145.632865:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as other-other

uk.search.yahoo.com - "mozilla firefox"

1.40.107

[24542:24542:0630/173353.184219:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.86

[24976:24976:0630/173525.163903:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as technology & computing-software

stephendonner · 2022-07-01T00:43:32Z

@tmancey can you check both @LaurenWags and my results, above?

In particular, that examples like search.yahoo.com and uk.search.yahoo.com both classify text -- we don't exclude classification due to locale, right?

tmancey · 2022-07-05T18:49:39Z

@stephendonner LGTM, all working as expected. The ML text classifier will decide if it can classify the text, but for the above this is all working as expected. Thanks

Uni-verse · 2022-07-05T19:11:00Z

Verified on Samsung GS 21 5G & Galaxy Tab S7 using

Brave	1.41.91 Chromium: 103.0.5060.114 (Official Build) beta (64-bit) 
Revision	a1c2360c5b02a6d4d6ab33796ad8a268a6128226-refs/branch-heads/5060@{#1124}
OS	Android 12; Build/SP1A.210812.016

STR from brave/brave-core#13522 (comment):

installed 1.41.91
launched Brave
Enabled ads, verbose logging for ads in QA pref.
loaded the base URL noted, for each
searched for the quoted term
checked console logs for text-classification snippet

amazon.com - "fishing gear"

Results

1.41.91
07-05 11:52:05.227 10630 10630 V chromium: [VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as sports-fishing

1.40.79
07-05 12:35:46.048 22798 22798 V chromium: [VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as sports-fishing

duckduck.com - "Top gun maverick"

Results

1.41.91
07-05 12:03:47.153 10630 10630 V chromium: [VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as arts & entertainment-film

1.40.79
07-05 12:42:24.008 22798 22798 V chromium: [VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

search.brave.com - "Obi wan kenobi"

Results

1.41.91
07-05 12:08:27.126 10630 10630 V chromium: [VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as hobbies & interests-sci-fi

1.40.79
07-05 12:45:05.436 22798 22798 V chromium: [VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

github.com - "cardano"

Results

1.41.91
07-05 12:10:29.702 10630 10630 V chromium: [VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as technology & computing-programming

1.40.79
07-05 12:49:09.538 22798 22798 V chromium: [VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

youtube.com - "crypto"

Results

1.41.91
07-05 12:17:25.017 10630 10630 V chromium: [VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as education-homeschooling

1.40.79
07-05 12:51:30.818 22798 22798 V chromium: [VERBOSE1:text_classification_processor.cc(65)] Classified text with the top segment as crypto-crypto

MadhaviSeelam · 2022-07-07T23:57:04Z

Verification PASSED using

Brave | 1.41.91 Chromium: 103.0.5060.114 (Official Build) beta (64-bit)
-- | --
Revision | a1c2360c5b02a6d4d6ab33796ad8a268a6128226-refs/branch-heads/5060@{#1124}
OS | Windows 11 Version 21H2 (Build 22000.739)

Verified test plan from brave/brave-core#13522 for sites as below:

installed 1.41.91
launched Brave
enabled ads
visited some of the sites as mentioned in the test plan
searched for a term
checked console logs for text-classification snippet

baidu.com - `wine`

1.40.113

[29400:15920:0707/150007.827:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.91

[13716:27124:0707/145308.881:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as food & drink-wine

qwant.com - `iPhone`

1.40.113

[9656:28920:0707/152523.268:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.91

[22368:6708:0707/164144.117:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as technology & computing-software

semanticscholar.org - `corona`

1.40.113

[15552:22112:0707/153502.686:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.91

[11944:29600:0707/164302.347:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as business-business

findx.com - `dogs`

1.40.113

[640:21512:0707/154903.703:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.91

[28400:16208:0707/164432.537:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as gaming-gaming

google.com - `toyota`

1.40.113

[21608:27868:0707/151942.766:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.91

[13340:21536:0707/164532.108:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as automotive-automotive

yandex.com - mail

1.40.113

[16300:26528:0707/164738.480:VERBOSE1:ads_impl.cc(219)] Search engine pages are not supported for text classification

1.41.91

[17688:9740:0707/164826.477:VERBOSE1:text_classification_processor.cc(64)] Classified text with the top segment as technology & computing-technology & computing

jsecretan added bug priority/P3 The next thing for us to work on. It'll ride the trains. feature/ads labels Sep 12, 2019

jsecretan assigned tmancey Sep 12, 2019

tmancey mentioned this issue Sep 23, 2019

Search engine result pages are not being classified brave/brave-core#3510

Closed

32 tasks

tmancey changed the title ~~SERP pages are not being classified in the ads user model~~ Fixes SERP pages are not being classified in the ads user model Sep 23, 2019

tmancey added the QA/Yes label Sep 23, 2019

tmancey added this to the 0.72.x - Nightly milestone Sep 23, 2019

jsecretan changed the title ~~Fixes SERP pages are not being classified in the ads user model~~ SERP pages are not being classified in the ads user model Oct 1, 2019

NejcZdovc modified the milestones: 0.72.x - Dev, 0.73.x - Nightly Oct 16, 2019

LaurenWags added the release-notes/include label Nov 20, 2019

NejcZdovc removed this from the 1.2.x - Dev milestone Nov 22, 2019

tmancey removed the release-notes/include label Jan 20, 2020

tmancey changed the title ~~SERP pages are not being classified in the ads user model~~ Search engine result pages are not being classified Mar 23, 2020

tmancey self-assigned this May 29, 2022

tmancey added priority/P3 The next thing for us to work on. It'll ride the trains. QA/Yes release-notes/exclude labels May 29, 2022

tmancey mentioned this issue May 29, 2022

Classify search engine results page content brave/brave-core#13522

Merged

25 tasks

tmancey added the OS/Android Fixes related to Android browser functionality label May 31, 2022

tmancey closed this as completed in brave/brave-core#13522 May 31, 2022

tmancey added this to the 1.41.x - Nightly milestone May 31, 2022

LaurenWags added the QA/Test-All-Platforms label Jun 21, 2022

LaurenWags added the QA/In-Progress Indicates that QA is currently in progress for that particular issue label Jun 29, 2022

LaurenWags added QA Pass-macOS and removed QA/In-Progress Indicates that QA is currently in progress for that particular issue labels Jun 29, 2022

LaurenWags mentioned this issue Jun 29, 2022

Optimize classification text handling (in brave ads) #23329

Closed

stephendonner added the QA Pass-Linux label Jul 1, 2022

Uni-verse added QA Pass - Android ARM QA Pass - Android Tab labels Jul 5, 2022

MadhaviSeelam added the QA Pass-Win64 label Jul 7, 2022

tmancey added this to Ads Jun 10, 2024

tmancey moved this to Done in Ads Jun 10, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Classify search engine results page content #6000

Classify search engine results page content #6000

jsecretan commented Sep 12, 2019 •

edited by tmancey

Loading

LaurenWags commented Jun 29, 2022 •

edited

Loading

stephendonner commented Jun 30, 2022 •

edited

Loading

stephendonner commented Jul 1, 2022

tmancey commented Jul 5, 2022

Uni-verse commented Jul 5, 2022 •

edited

Loading

MadhaviSeelam commented Jul 7, 2022

Classify search engine results page content #6000

Classify search engine results page content #6000

Comments

jsecretan commented Sep 12, 2019 • edited by tmancey Loading

LaurenWags commented Jun 29, 2022 • edited Loading

stephendonner commented Jun 30, 2022 • edited Loading

Verification PASSED using

Shared Steps:

duckduckgo.com - "shoes"

search.brave.com - "barbie dolls"

stackoverflow.com - "python"

twitter.com/explore - "covid"

excite.com - "ukraine"

search.yahoo.com - "news"

uk.search.yahoo.com - "mozilla firefox"

stephendonner commented Jul 1, 2022

tmancey commented Jul 5, 2022

Uni-verse commented Jul 5, 2022 • edited Loading

MadhaviSeelam commented Jul 7, 2022

baidu.com - wine

qwant.com - iPhone

semanticscholar.org - corona

findx.com - dogs

google.com - toyota

yandex.com - mail

jsecretan commented Sep 12, 2019 •

edited by tmancey

Loading

LaurenWags commented Jun 29, 2022 •

edited

Loading

stephendonner commented Jun 30, 2022 •

edited

Loading

Verification `PASSED` using

Uni-verse commented Jul 5, 2022 •

edited

Loading

baidu.com - `wine`

qwant.com - `iPhone`

semanticscholar.org - `corona`

findx.com - `dogs`

google.com - `toyota`