Page MenuHomePhabricator
Create Task
Maniphest T260564

Deliver 4 milestones towards 1 model to understand the diffusion of content on the Wikimedia projects
Closed, ResolvedPublic
Actions

Description

This task reunites our efforts in modeling content diffusion in Wikipedia.

Goals by Quarter:

  • Q1 Milestone 1: Prepare the datasets for training 2 models about diffusion of content. (T252933)
  • Q2 Milestone 2: Train and validate 1 model: predict the likelihood of an item spreading from a given project to another; (T252933]
  • Q3 Milestone 3: Train and validate a second model to predict the likelihood of a project receiving information from another project.
  • Q4 Milestone 4: Publish the work and share insights from the research with the Wikimedia community.

Links to:
Paper: https://arxiv.org/abs/2103.16613
Code: https://github.com/rodolfovalentim/wikipedia-content-propagation
Meta: https://meta.wikimedia.org/wiki/Research:Exploration_on_content_propagation_across_Wikimedia_projects
Dataset: https://zenodo.org/record/4433137

Related Objects
Search...

Event Timeline

diego created this task.Aug 17 2020, 3:13 PM
Restricted Application added a subscriber: Aklapper. · View Herald TranscriptAug 17 2020, 3:13 PM
diego triaged this task as High priority.Aug 17 2020, 3:13 PM
diego added a subtask: T252933: Exploration on content propagation across Wikimedia projects..
diego added a comment.Sep 4 2020, 8:47 PM

Updates

  • The two datasets have been prepared:
    • One dataset with items that propagates across Wikipedias, removing bot activity.
    • Another dataset about external references (links) across projects.
diego updated the task description. (Show Details)Sep 4 2020, 8:47 PM
leila moved this task from FY2020-21-Research-July-September to FY2020-21-Research-October-December on the Research board.Oct 16 2020, 8:35 PM
leila edited projects, added: Research (FY2020-21-Research-October-December); removed: Research (FY2020-21-Research-July-September).
diego added a comment.Nov 16 2020, 11:31 AM

Updates

  • The dataset are ready. We are waiting for the paper to be published to share the link in public. Currently, the datasets are available under request via email.
diego updated the task description. (Show Details)Dec 11 2020, 2:44 PM
diego added a comment.Dec 11 2020, 2:46 PM

Updates

  • Depending on the results for the paper submission (received border line evaluation), we are planning to publish the dataset separately from the model. In the case of publishing the dataset separately, this will be done during Q3.
leila moved this task from FY2020-21-Research-October-December to FY2020-21-Research-January-March on the Research board.Jan 12 2021, 6:57 PM
leila edited projects, added: Research (FY2020-21-Research-January-March); removed: Research (FY2020-21-Research-October-December).
diego added a comment.Jan 15 2021, 5:53 PM

Updates

  • We submitted the propagation dataset to ICWSM.
  • We are building a new model considering content popularity (pageviews).
diego added a comment.Jan 23 2021, 1:49 AM

Updates

  • The dataset to model content propagation has been published in Zenodo.
diego mentioned this in T274400: Request creation of research-collaborations-api VPS project.Feb 10 2021, 6:49 PM
diego added a comment.Feb 19 2021, 7:02 PM

Updates

  • New experiments using pageviews as feature to predict content propagation.
diego added a comment.Feb 26 2021, 9:30 PM

Updates

  • Pageviews added an improvement over %5 on predicting content propagation.
  • We are currently working on add content-related information (articles meta-data) to de model. This will allow to study the effects content quality on the spread patterns.
diego added a comment.Mar 13 2021, 2:42 AM

Updates

  • We are improvements on the model to predict the next language of propagation.
  • We are working on modeling changes of content propagation behavior depending on the article reliability.
diego added a comment.Mar 19 2021, 7:34 PM

Update

  • The paper with the dataset and first model has been accepted on ICWSM'21
  • We are working on the camera ready version.
diego added a comment.Mar 26 2021, 11:20 PM

Update

  • Finishing the camera ready version for ICWSM'21 paper.
  • We have trained a new model, content information, to predict the likelihood of item to propagate to other projects.
  • We are working on documenting our new results.
diego updated the task description. (Show Details)Mar 26 2021, 11:20 PM
diego added a comment.Apr 3 2021, 12:51 AM

Update

leila moved this task from FY2020-21-Research-January-March to FY2020-21-Research-April-June on the Research board.Apr 29 2021, 1:01 AM
leila edited projects, added: Research (FY2020-21-Research-April-June); removed: Research (FY2020-21-Research-January-March).
diego added a comment.Apr 30 2021, 9:59 PM

*Update*

  • Preparing the ICWSM presentation.
diego added a comment.May 7 2021, 7:18 PM

Update

diego added a comment.May 14 2021, 3:02 PM

Updates

  • The work is published, poster and presentation submitted.
diego closed subtask T252933: Exploration on content propagation across Wikimedia projects. as Resolved.May 14 2021, 3:08 PM
diego added a comment.May 21 2021, 10:06 PM

Updates

  • No updates.
diego added a comment.Jun 4 2021, 6:44 PM

Updates

  • We will be presenting this work at ICSWM, on June 10th.
diego added a comment.Jun 11 2021, 6:31 PM

Updates

  • The paper was successfully presented at ICWSM'21.
  • We also shared this and the "Wiki-Reliability" paper on the wiki-research mailing list.

With these last two updates, we have finalized this task.

diego closed this task as Resolved.Jun 11 2021, 6:35 PM
diego updated the task description. (Show Details)
Content licensed under Creative Commons Attribution-ShareAlike (CC BY-SA) 4.0 unless otherwise noted; code licensed under GNU General Public License (GPL) 2.0 or later and other open source licenses. By using this site, you agree to the Terms of Use, Privacy Policy, and Code of Conduct. · Wikimedia Foundation · Privacy Policy · Code of Conduct · Terms of Use · Disclaimer · CC-BY-SA · GPL · Credits