In T235309#5805098, @Nuria wrote:

We will met with @JFishback_WMF and let you know of next steps

Weekly update: paper was accepted with good reviews !! there is some work to do to respond to reviewers' comments, which we will do in the next month.

Weekly update: sent invitation emails to a few speakers and upcoming submission deadline reminder.

Weekly update:

• downloaded the list of coco-stuff classes which include highly generic categories of people, animals, and things which exist in the visual world: https://github.com/nightrome/cocostuff
• downloaded the list of categories in Commons, with the counts of the number of images per categories.
• to create the initial seed of categories we want to consider for object categorization in Commons, I computed fasttext vectors on both COCO categories and Commons Categories, and I am checking what are the commons categories that we can use to represent COCO categories.

Weekly update: progress as expected. I had to double check the output of the citation needed models as there were some inconsistencies, which were solved by looking at some of the data. There is need for a text pre-processing pipeline which is close to the one we used for training, otherwise the model might give some unexpected output.

Weekly Update: none. Team work is on hold until ICWSM deadline.

Weekly update: project proposal is ready. Investigating possibilities with Rutgers' university. We might decide to merge collaboration with Rutgers' within the Visual Knowledge Gaps project (T242603). In that case, I will reach out to further contacts starting last week of Jan .

Visualization available at http://wikiview.net

Done, presentation "Machine Vision in Social Computing" available at: https://docs.google.com/presentation/d/1O1Ni40tovcFIh_Btj3qpWFh5Pz8j5Bh_U0u5rqfTbW0/edit?usp=sharing

The output of the first stage of this research, as well as the opportunities arisen for collaborations and internships, led to the decision of having an intermediate step for this task.
In collaboration with @AikoChou, an Outreachy intern, we will have a system that can release periodic data dumps exposing sentences in Wikipedia needing citations. This task is tracked at: T233707.

I'll leave this task open until we hear back from the Web Conference

When started this sub-task as part of our efforts on citation usage studies, we reealised that this is a huge effort that will be spun off to a completely new project

@leila @Isaac does this look good? https://commons.wikimedia.org/wiki/File:Wiki_Workshop_2020_banner.png

Update after changing learning rate and modifying few parameters in the training:

• we reach 91% accuracy on the validation set (2% improvement compared to previous model) and a significant loss reduction.
• Manual validation on an external dataset shows improvements on overfitting (80% of new photos and 100% of new graphics are correctly recognized)
• Problems persist for near-abstract photos and for extremely low-res photos
• All this with an training time of 18-minutes only for a dataset of 150k images!

In T237605#5700591, @elukey wrote:

In T237605#5700493, @Miriam wrote:

Hi! Can I have access too please? My username is mirrys

elukey@krb1001:~$ sudo manage_principals.py create mirrys --email_address=mredi@wikimedia.org
Principal successfully created.
Successfully sent email to mredi@wikimedia.org

Done! :)

Hi! Can I have access too please? My username is mirrys

Finally I managed to do some progress. I built a simple model for testing purposes.

Proposal submitted on September 30th: https://github.com/OlehOnyshchak/WikiImageRecommendation/blob/master/Papers/Project_Proposal/paper.pdf

In T234519#5570276, @Moniquedejesus wrote:

Hello, @Miriam! My name is Monique and I don't know python, it will be difficult for me? Thank you in advance!

In T234606#5549527, @Achillesheel02 wrote:

Hi @Samwalton9 ,
Which section title takes precedence? I noticed there's different levels and was wondering about which one to use.

@Achillesheel02 thanks for the feedback and yes, please send me to my email.
Thanks!

In T234519#5543837, @Shamima19 wrote:

Hi @Miriam ..can we use python 3.6 for the tasks?

In T234519#5543780, @Kabarozi wrote:

Also, am a little confused if this is part of the projects for interns,i saw Outreachy (round 9) and jumped right in , kindly guide me through

In T234519#5543781, @Pgadige01 wrote:

Thank you for creating this task, @Miriam.
I was able to install the required libraries in my development environment.
I tried looking for sample_data.txt file in the project's GitHub repo, but I'm afraid I can't seem to find it there. The only data file that I can see is test_input_data_sample.txt file.
I'll appreciate a hint on locating this sample_data.txt file.

Hi @Lucideuclid - many thanks for your interest in this project :)

Hi @Nuria and @elukey - please feel free to drop this dataset from mysql. @tizianopiccardi -the main user for this data - also confirmed.

Thanks @elukey! @MGerlach you should now be able to access the notebooks. Thanks!

Hi @elukey -our new team member, @MGerlach, is having the same issue (he is not able to access SWAP)- see T232707#5502540
Could you please add him to the wmf group as well?

Examples of issues of Google Vision API, related to the fact that their models are probably trained on unevenly distributed data: https://docs.google.com/presentation/d/1qfD3q9Ij79_luAKXNdVvZoPFsUvFQKcvIGmIvWC19z0/edit?usp=sharing

So Rachel said we are welcome to join the open mic session at the beginning of the hackathon. Each speaker has 1-1.5 minutes to pitch projects " that newcomers can participate in and/or needs significant collaboration from other attendees".
She suggested that, once we have the mini-projects ready, we share them on phab : https://phabricator.wikimedia.org/project/view/3922 and also on the Wikimania Hackathon telegram channel.

@leila mini-projects for hackathon can be added to the phab workboard for the Hackathon: https://phabricator.wikimedia.org/project/view/3922/. Checking now with Rachel about the pitches.