Page MenuHomePhabricator

Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all)
Open, MediumPublic

Description

At the moment, we are using LTR on large wikis, where we have enough data to train models. By increasing data retention to > 90 days (which requires ensuring that data is anonymized), we could apply the same techniques to more wikis. LTR is likely to increase the quality of search, and to continue to increase it over time.

Event Timeline

@Gehel: Does Q2 refer to WMF's Financial Year 2019-20? If so, please say so for the sake of transparency and not to collide with other years. Also see https://www.mediawiki.org/wiki/Phabricator/Project_management#Use_plain_language,_define_actions_and_expected_results - thanks!

Gehel renamed this task from [Objective Q2] Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all) to [Objective Q2] (6) Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all).Nov 8 2019, 3:55 PM
Gehel renamed this task from [Objective Q2] (6) Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all) to [Objective Fiscal 19-20/Q2] (6) Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all).Nov 8 2019, 3:59 PM
CBogen renamed this task from [Objective Fiscal 19-20/Q2] (6) Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all) to Learning to Rank (LTR) applied to additional languages and projects to improve ranking (needs experimentation, might not work at all).Sep 23 2020, 2:22 PM
CBogen triaged this task as Medium priority.