AI

ai@lists.wikimedia.org

1 participants
265 discussions

Participate in a labeling campaign about verifiability of unsourced statements!
by Jonathan Morgan 19 Aug '18

19 Aug '18

Good morning/afternoon/evening everyone, If you are an editor of the French, Italian or English Wikipedia, and you are curious about how to contribute to technologies for improving verifiability of Wikipedia articles, please read on—we need your help! In the context of the Knowledge integrity <https://meta.wikimedia.org/wiki/Knowledge_Integrity> program, we (the WMF Research team <http://research.wikimedia.org>) are studying ways to flag unsourced statements needing a citation <https://meta.wikimedia.org/wiki/Research:Identification_of_Unsourced_Statem…> using machine learning, with the aim of identifying areas where adding high quality citations is particularly urgent or important. Following the success of the first labeling campaign <https://meta.wikimedia.org/wiki/Research:Identification_of_Unsourced_Statem…>, we now need to collect additional, high-quality labeled data regarding why sentences need citations. You are invited to participate in a second annotation task <https://meta.wikimedia.org/wiki/Research:Identification_of_Unsourced_Statem…>. We used your input from the last experiment to generate a taxonomy of reasons <https://meta.wikimedia.org/wiki/Research:Identification_of_Unsourced_Statem…> why editors add citations. With this taxonomy now embedded in the interface, the annotation experience will be much faster and fun. If you are interested in participating, please go to http://labels.wmflabs.org/ui/enwiki/ (replace enwiki with itwki or frwiki if you speak Italian or French), login, and from *'**Labeling Unsourced Statements II’**,* request one (or more) workset. For each task in a workset, the tool will show you an unsourced sentence in an article and ask you to annotate it. You can then label the sentence as needing an inline citation or not, and specify a reason for your choice from a drop-down menu. If you can't respond please select 'skip'. You can also sign up by (optionally) adding your name on this page <https://meta.wikimedia.org/wiki/Research:Identification_of_Unsourced_Statem…> to receive updates about future campaigns and results from this research If you have any question/comment on this project, please let us know by contacting miriam(a)wikimedia.org or leaving a message on the talk page of the project <https://meta.wikimedia.org/wiki/Research_talk:Identification_of_Unsourced_S…>. Thank you for your time! Miriam, Jonathan, and Dario -- Jonathan T. Morgan Senior Design Researcher Wikimedia Foundation User:Jmorgan (WMF) <https://meta.wikimedia.org/wiki/User:Jmorgan_(WMF)>

1 0

Fwd: [Analytics] Wikimedia Research Showcase August 13 2018 at 11:30 AM (PDT) 18:30 UTC
by Pine W 11 Aug '18

11 Aug '18

Forwarding. Pine ( https://meta.wikimedia.org/wiki/User:Pine ) ---------- Forwarded message --------- From: Sarah R <srodlund(a)wikimedia.org> Date: Fri, Aug 10, 2018 at 10:46 PM Subject: [Analytics] Wikimedia Research Showcase August 13 2018 at 11:30 AM (PDT) 18:30 UTC To: <wikimedia-l(a)lists.wikimedia.org>, <wiki-research-l(a)lists.wikimedia.org>, <analytics(a)lists.wikimedia.org> Hi Everyone, The next Wikimedia Research Showcase will be live-streamed Wednesday, August 13 2018 at 11:30 AM (PDT) 18:30 UTC. YouTube stream: https://www.youtube.com/watch?v=OGPMS4YGDMk As usual, you can join the conversation on IRC at #wikimedia-research. And, you can watch our past research showcases here. <https://www.mediawiki.org/wiki/Wikimedia_Research/Showcase#Upcoming_Showcase> Hope to see you there! This month's presentations is: *Quicksilver: Training an ML system to generate draft Wikipedia articles and Wikidata entries simultaneously* John Bohannon and Vedant Dharnidharka, Primer The automatic generation and updating of Wikipedia articles is usually approached as a multi-document summarization task: Given a set of source documents containing information about an entity, summarize the entity. Purely sequence-to-sequence neural models can pull that off, but getting enough data to train them is a challenge. Wikipedia articles and their reference documents can be used for training, as was recently done <https://arxiv.org/abs/1801.10198> by a team at Google AI. But how do you find new source documents for new entities? And besides having humans read all of the source documents, how do you fact-check the output? What is needed is a self-updating knowledge base that learns jointly with a summarization model, keeping track of data provenance. Lucky for us, the world’s most comprehensive public encyclopedia is tightly coupled with Wikidata, the world’s most comprehensive public knowledge base. We have built a system called Quicksilver uses them both. _______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

2 1

Using AI to identify counterfeit images
by Pine W 26 Jun '18

26 Jun '18

This is in today's news: https://www.dpreview.com/news/5039410137/adobe-ai-spots-tampered-images-by-… Pine ( https://meta.wikimedia.org/wiki/User:Pine )

1 0

Idle GPU servers - 1080TI
by Pavel Machalek 06 Jun '18

06 Jun '18

Dear AI ers, we recently finished a large Tensorflow project and our two servers are sitting idle. Let me know if you want to rent machines with TEN GeForce GTX 1080TIs, and over 250GB RAM and 48 CPUs. Latest CUDA and NVIDIA drivers installed. Cheers Paul

2 1

Scoring platform status update
by Adam Wight 02 May '18

02 May '18

We've summarized our technical work over the last three months, please see the blog post for details: https://phabricator.wikimedia.org/phame/post/view/104/status_update_may_2_2… Our novel, collaborative auditing system "judgment and dialogue engine" (JADE) is deployed to the beta cluster and is now available for tool developers to explore and integrate with. Read more here, https://mediawiki.org/wiki/JADE A dynamically updated table of ORES support is now available for anyone tracking our progress: https://tools.wmflabs.org/ores-support-checklist/ Regards, Adam

1 0

Study on music habits
by Sylvain Castagnos 26 Apr '18

26 Apr '18

*** Apologies for cross postings *** Hello, We are researchers at the LORIA laboratory (Nancy, France) and we are conducting a study on musical listening habits of the population over the world. Our goal is to better understand the preferences, expectations, and needs of users as regards music. This questionnaire is divided into 3 sections and its duration is estimated at 5 minutes. Could you help us by completing this form and spreading it widely? Fill out form <https://docs.google.com/forms/d/e/1FAIpQLSeWqSXQh0D872L6cfCEcHhB1ojyBUntsrz…> For any questions related to this experience, please contact us at the following email address: etude-kiwi(a)loria.fr. The answers to this questionnaire are collected anonymously. Thank you for your participation ! Best regards, -- Sylvain CASTAGNOS KIWI Team LORIA Laboratory

1 0

In the news: "Your next phone may have an ARM machine learning processor"
by Pine W 18 Feb '18

18 Feb '18

In case you're interested: https://techcrunch.com/2018/02/13/your-next-phone-may- have-an-arm-machine-learning-processor/ Pine <https://meta.wikimedia.org/wiki/User:Pine> <https://en.wikipedia.org/wiki/User:CatherineMunro/Bright_Places>

1 0

Re: [AI] [Wikimedia-l] Siamese networks and image classification
by Pine W 31 Jan '18

31 Jan '18

Hi John, I think that these questions might be best asked on the AI mailing list, so I'm copying this thread to that list. I would suggest that further discussion should take place on the AI mailing list, because that's where the WMF experts on AI subjects are likely to participate. Thanks for your interest in these questions. I too would be interested in automating image categorization for Commons. I think that full automation of image categorization would be a long term project, perhaps over the course of decades, but there may be ways to make incremental progress. :) Pine <https://meta.wikimedia.org/wiki/User:Pine> <https://en.wikipedia.org/wiki/User:CatherineMunro/Bright_Places> On Wed, Jan 24, 2018 at 1:20 PM, John Erling Blad <jeblad(a)gmail.com> wrote: > I had a plan to do a more thorough description on meta, but my plans are > always to optimistic… ;) > > Question 1: I wonder if anyone has done any work on categorization of > images on Commons by using neural nets. > Question 2: I wonder if anyone has used Siamese networks to do more general > categorization of images. > > A Siamese network is a kind of network used for face recognition. It is > called Siamese network because it is two parallel networks in its simplest > form, but one of the networks are precomputed so in effect you only > evaluates one of the networks. Usually you have one face provided by a > camera and run through one network, and a number of other precomputed faces > for the other network. When the difference is small, then you have a face > match. > > In general you can have any kind of image as long as you can train a kind > of fingerprint type of vector. This fingerprint will then acts like > locality-sensitive hashing [1]. Because you can create this LSH from the > vector, you can store alternate models in database, so you can later search > it and find usable models. Those models can then interpret the vector form > of the fingerprint, and by evaluating those models with actual input (ie > the fingerprint vector) you can get an estimate for how probable it is that > the image belong to a specific category. > > What you would get from the previous process is a list of probable > categorizations. That list can be sorted, and a user trying to categorize > an image can then chose to select some of the categories. > > Note that the output of the network is not just a few categories, it can be > a variant number of models that outputs a probability for a specific > category. It is a bit like a map-reduce, where the you find (filter) the > possible models and then evaluate (map) those models with the fingerprint > vector. > > It is perhaps not obvious, but the idea behind Siamese networks are two > parallel networks, but the implementation is with a single active network. > Also the implemented network has a first part that computes the fingerprint > vector (kind of a bottleneck network) and the second part are the stored > models that takes this fingerprint vector and calculates a single output > for the probability of a single category. > > [1] https://en.wikipedia.org/wiki/Locality-sensitive_hashing > > On Mon, Jan 22, 2018 at 6:47 AM, Pine W <wiki.pine(a)gmail.com> wrote: > > > Hi John, > > > > I am having a little trouble with understanding your email from January > > 15th. Could you perhaps state your question or point in a different way? > > > > Pine <https://meta.wikimedia.org/wiki/User:Pine> > > <https://en.wikipedia.org/wiki/User:CatherineMunro/Bright_Places> > > > > On Mon, Jan 15, 2018 at 5:55 AM, John Erling Blad <jeblad(a)gmail.com> > > wrote: > > > > > This is the same as the entry on the wishlist for 2016, but describes > the > > > actual method. > > > https://meta.wikimedia.org/wiki/2016_Community_Wishlist_ > > > Survey/Categories/Commons#Use_computer_vision_to_propose_categories > > > > > > Both contrastive and triplet loss can be used while learning, but > neither > > > are described at Wikipedia. > > > > > > On Sun, Jan 14, 2018 at 8:16 PM, Pine W <wiki.pine(a)gmail.com> wrote: > > > > > > > Hi John, > > > > > > > > I have not heard of an initiative to use Siamese neural networks for > > > image > > > > classifications on on Commons. You might make a suggestion on the AI, > > > > Research, and/or Commons mailing lists regarding this idea. You might > > > also > > > > make a suggestion in IdeaLab > > > > <https://meta.wikimedia.org/wiki/Grants:IdeaLab>. > > > > > > > > Pine <https://meta.wikimedia.org/wiki/User:Pine> > > > > <https://en.wikipedia.org/wiki/User:CatherineMunro/Bright_Places> > > > > > > > > On Sun, Jan 14, 2018 at 3:46 AM, John Erling Blad <jeblad(a)gmail.com> > > > > wrote: > > > > > > > > > Has anyone tried to use a Siamese neural network for image > > > classification > > > > > at Commons? I don't know if it will be good enough to run in > > autonomous > > > > > mode, but it will probably be a huge help for those that do manual > > > > > classification. > > > > > > > > > > Imagine a network providing a list of possible categories, and the > > user > > > > > just ticks off usable categories. > > > > > > > > > > A Siamese network can be learned by using a triplet loss function, > > > where > > > > > the anchor and the positive candidate comes from the same category, > > and > > > > the > > > > > negative candidate comes from an other category but are otherwise > > close > > > > to > > > > > the anchor. > > > > > > > > > > Output from the network is like a fingerprint, and those > fingerprints > > > can > > > > > be compared to other images with known fingerprints, or against a > > > > > generalized fingerprint for a category. > > > > > > > > > > John Erling Blad >

1 0

Scoring platform team status update
by Adam Wight 30 Jan '18

30 Jan '18

Friends, Anyone interested in ORES and Scoring Platform’s recent progress can read an exhaustive list on our engineering blog, https://phabricator.wikimedia.org/phame/post/view/84/status_update_january_… <https://phabricator.wikimedia.org/phame/post/view/84/status_update_january_…> Aside from supporting several new languages, the biggest changes that might affect other developers are that: we now supply enwiki new article draft quality predictions, already loaded into the MediaWiki database via Extension:ORES; and all of our models now include statistics that can be used to calculate appropriate cutoff points. Enjoy! -Amir, Aaron, Adam, and Sumit

1 0

Research Project on Real-Time Vandal Detection during Editing
by Martin Potthast 20 Jan '18

20 Jan '18

Hi everyone, we [1] would like to announce a research project with the goal of studying whether user interactions recorded at the time of editing are suitable to predict vandalism in real time. Should vandal editing behavior be sufficiently different from normal editing behavior, this would allow for a number of interesting real-time prevention techniques. For example: - withholding confidently suspicious edits for review before publishing them, - a popup asking "I am not a vandal" (as in Google's "I am not a robot") to analyze vandal reactions, - a popup with a chat box to personally engage vandals, e.g., to help them find other ways of stress relief or to understand them better, - or at the very least: a new signal to improve traditional vandalism detectors. We have set up a laboratory environment to study editor behavior in a realistic setting using a private mirror of Wikipedia. No editing whatsoever is conducted on the real Wikipedia as part of our experiments, and all test subjects of our user studies are made aware of the experimental nature of their editing. We plan on making use of crowdsourcing as a means to attain scale and diversity. If you wish to participate in this study as a test subject yourself, please get in touch. The more diversity, the more insightful the results will be. We are also happy to collaborate and to answer all questions that may arise in relation to the project. For example, our setup and tooling may turn out to be useful to study other user behavior-related things without having to actually deploy experiments within the live MediaWiki. Best, Martin PS: The AICaptcha project seems most closely related. @Vinitha and Gergő: If you wish, we can set up a Skype meeting to talk about a avenues for collaboration. [1] A group of students and researchers from Bauhaus-Universität Weimar ( www.webis.de) and Leipzig University (www.temir.org); project PI: Martin Potthast.

1 0

← Newer
1
...
10
11
12
13
14
15
16
...
27
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

AI