"use strict";(self.webpackChunk_toloka_portal_gatsby_renderer=self.webpackChunk_toloka_portal_gatsby_renderer||[]).push([[3769],{34151:function(e,t,n){n.d(t,{r:function(){return _}});var r=n(4942),a=n(37990),i=n(93252),s=n(52893),o=n(74533),l=n(88586),c=n(88709),d=n(51512),u=n(92401),f=n(44172),p=n(25109),m=n(18987),h=n(13598),g=n(62769),x=n(67294),j=n(3468),b=n(81547),v=n(85893);function w(e,t){var n=Object.keys(e);if(Object.getOwnPropertySymbols){var r=Object.getOwnPropertySymbols(e);t&&(r=r.filter((function(t){return Object.getOwnPropertyDescriptor(e,t).enumerable}))),n.push.apply(n,r)}return n}function k(e){for(var t=1;tt.date?1:-1})).slice(0,4)}),[_]);return(0,v.jsxs)(v.Fragment,{children:[(0,v.jsx)(s.q,{og:{img:"https://tlk-infra-front.azureedge.net/portal-static/images/page-head/eventsOg/"+t+".jpg"}}),(0,v.jsx)(o.n,{}),(0,v.jsxs)(l.i,{headerPadding:!0,children:[(0,v.jsx)(a.h,{defaultBackground:"dark"}),w,(D||C&&+new Date(C)>R)&&!1===t.startsWith("watch")&&I?(0,v.jsxs)(v.Fragment,{children:[(0,v.jsx)(d.w,{offset:40}),(0,v.jsx)(p.L,{header:{title:T,info:O},theme:"sandSection",id:"registration_form",children:(0,v.jsx)(m.G,{id:I,slug:P||(b.j?window.location.href+"?registered=true":""),leadSource:"Event Registration (website)",textareaPlaceholder:F,textareaRequired:S})})]}):null,(0,v.jsx)(d.w,{offset:40}),V.length>0?(0,v.jsx)(p.L,{header:{title:"Similar "+(0,g.s)(n)+"s"},children:(0,v.jsx)(f.X,{events:V.map((function(e){return e}))})}):null,(0,v.jsx)(u.F,{leftColumn:{title:{text:"Don't miss"},text:(0,v.jsxs)(v.Fragment,{children:["Be sure to attend our informative workshops,"," ",(0,v.jsx)(j.p.Desktop,{children:(0,v.jsx)("br",{})}),"tutorials, and webinars."]})},rightColumn:{input:{label:"Subscribe to updates",placeholder:"Your e-mail"},button:{text:"Get updates",view:"action",type:"submit",size:"l"},successText:"All set! You’ll get an email from us shortly — just click the link to confirm. If it doesn’t show up, check your spam folder.",sendingData:{address:"/api/sender/subscribe",listId:"8AR2Z6Y3-R251"}}}),(0,v.jsx)(d.w,{offset:40}),(0,v.jsx)(i.$,{})]}),(0,v.jsx)(c.f,{})]})}},94501:function(e,t,n){n.d(t,{B:function(){return S}});var r=n(67294),a="bl_Z",i="bl_c7",s="bl_x",o="bl_w",l=n(79615),c=n(85893),d=function(){return(0,c.jsx)("svg",{xmlns:"http://www.w3.org/2000/svg",width:"12",height:"13",fill:"none",children:(0,c.jsx)("path",{fillRule:"evenodd",clipRule:"evenodd",d:"M1.62.4C.72.4 0 1.13 0 2c0 .87.72 1.58 1.62 1.58A1.6 1.6 0 0 0 3.23 2 1.6 1.6 0 0 0 1.62.4Zm.96 3.62H.19a.2.2 0 0 0-.19.2v7.75c0 .1.09.2.2.2h2.38c.1 0 .19-.1.19-.2V4.22a.2.2 0 0 0-.2-.2Zm4.25.87a2.83 2.83 0 0 1 2.1-.87c2.7 0 3.07 1.92 3.07 3.8v4.16a.2.2 0 0 1-.2.19H9.43a.2.2 0 0 1-.2-.2V8.3c0-1.03-.07-1.82-1.05-1.82-.9 0-1.24.5-1.24 1.76v3.75a.2.2 0 0 1-.2.19H4.36a.2.2 0 0 1-.2-.2V4.4c0-.1.09-.19.2-.19h2.3c.1 0 .18.09.18.2v.48Z",fill:"#fff"})})},u=n(4511),f=n(46237),p="dX_bk",m="dX_bl",h="dX_r",g="dX_bp",x="dX_q",j="dX_bm",b="dX_bn",v="dX_j",w="dX_g",k=function(e){var t=e.imageSrc,n=e.name,r=e.company,a=e.info,i=e.link;return(0,c.jsxs)("div",{className:p,children:[(0,c.jsx)("div",{className:m,children:(0,c.jsx)(l.P,{version:"v1",sizes:[560],imageId:t,className:h})}),(0,c.jsxs)("div",{className:g,children:[(0,c.jsx)(f.D,{className:x,level:5,children:n}),(0,c.jsx)("span",{className:j,children:r}),(0,c.jsx)("span",{className:b,children:a}),(0,c.jsxs)(u.r,{href:i.href,iconPosition:i.iconPosition,target:"_blank",className:v,children:[(0,c.jsx)("div",{className:w,children:(0,c.jsx)(d,{})}),i.text]})]})]})},y="c9_by",_="c9_h",D="c9_g2",I="c9_g3",P="c9_bl",F=function(e){var t=e.text,n=e.imageSrc,r=e.name,a=e.company,i=e.info,s=e.link;return(0,c.jsxs)(c.Fragment,{children:[(0,c.jsx)("div",{className:y}),(0,c.jsxs)("div",{children:[(0,c.jsx)("p",{className:_,children:t}),n&&r&&a&&i&&s?(0,c.jsxs)("button",{className:D,children:[(0,c.jsx)("div",{className:P,children:(0,c.jsx)(l.P,{version:"v1",sizes:[560],imageId:n})}),(0,c.jsx)("span",{children:r}),(0,c.jsx)("div",{className:I,children:(0,c.jsx)(k,{name:r,imageSrc:n,company:a,info:i,link:s})})]}):null]})]})},S=function(e){var t=e.content;return(0,c.jsx)("div",{className:a,children:(0,c.jsx)("div",{className:i,children:t.map((function(e,t){return(0,c.jsxs)(r.Fragment,{children:[(0,c.jsx)("div",{className:s,children:(0,c.jsx)("div",{children:e.date})}),(0,c.jsx)("div",{className:o,children:(0,c.jsx)(F,{text:e.text,imageSrc:e.imageSrc,name:e.name,company:e.company,info:e.info,link:e.link})})]},t)}))})})}},86784:function(e,t,n){n.d(t,{_:function(){return c}});var r=n(84401),a="b6_dx",i="b6_q",s="b6_fq",o=n(85893),l=function(e){var t=e.title,n=e.link;return(0,o.jsxs)("div",{className:a,children:[(0,o.jsx)("div",{className:i,children:t}),(0,o.jsx)("div",{children:(0,o.jsx)(r.z,{size:"m",themeColor:"light",type:"link",url:n,view:"pseudo",children:"Download"})})]})},c=function(e){var t=e.files;return(0,o.jsx)("div",{className:s,children:t.map((function(e){return(0,o.jsx)(l,{title:e.title,link:e.link},e.link)}))})}},18987:function(e,t,n){n.d(t,{G:function(){return g}});var r=n(84181),a=n(79700),i=n.n(a),s=n(84401),o=n(3468),l=n(29005),c=n(34305),d=n(54269),u=n(4511),f=n(67294),p=n(96268),m=n(29499),h=n(85893),g=function(e){var t=e.slug,n=e.id,a=e.isTestForm,g=e.leadSource,x=void 0===g?"WebsiteRegistrationForm":g,j=e.textareaRequired,b=void 0!==j&&j,v=e.textareaPlaceholder,w=void 0===v?"Please drop a few lines about your project":v,k=e.companyRequired,y=void 0===k||k,_=e.jobTitleRequired,D=void 0===_||_,I=e.additionalInputs,P=e.withoutTextArea,F=e.submitButtonText,S=void 0===F?"Submit":F,N=(0,m.useLocation)();(0,f.useEffect)((function(){N.search&&document.querySelectorAll("[id^='utm']").forEach((function(e){e&&(e.value=(0,p.t)(e.id)||"")}))}),[N.search]);var T=(0,f.useMemo)((function(){try{return new URLSearchParams(N.search).get("referer")||void 0}catch(e){return}}),[N.search]);return(0,h.jsxs)("form",{action:"https://webto.salesforce.com/servlet/servlet.WebToLead?encoding=UTF-8",method:"POST",children:[(0,h.jsx)("input",{type:"hidden",name:"oid",defaultValue:"00D7Q000001MbTr"}),(0,h.jsx)("input",{type:"hidden",name:"retURL",defaultValue:t.startsWith("http")?t:"https://toloka.ai/"+t}),(0,h.jsx)("input",{type:"hidden",name:"Campaign_ID",value:n}),(0,h.jsx)("input",{type:"hidden",name:"recordType",id:"recordType",value:"0127Q0000002WpG"}),(0,h.jsx)("input",{id:"00N7Q00000EpH4g",name:"00N7Q00000EpH4g",type:"hidden",value:T}),N.search?(0,h.jsxs)(h.Fragment,{children:[(0,h.jsx)("input",{id:"utm_source",name:"00N7Q000005OX3o",type:"hidden"}),(0,h.jsx)("input",{id:"utm_medium",name:"00N7Q000005OX3j",type:"hidden"}),(0,h.jsx)("input",{id:"utm_term",name:"00N7Q000005OX43",type:"hidden"}),(0,h.jsx)("input",{id:"utm_campaign",name:"00N7Q000005OX3t",type:"hidden"}),(0,h.jsx)("input",{id:"utm_content",name:"00N7Q000005OX3y",type:"hidden"})]}):null,(0,h.jsxs)("div",{className:i()(r.Z.inputsWrap,"light"),children:[(0,h.jsx)(l.I,{placeholder:"First name*",name:"first_name",required:!0}),(0,h.jsx)(l.I,{placeholder:"Last name*",name:"last_name",required:!0}),(0,h.jsx)(l.I,{placeholder:y?"Company*":"Company",name:"company",required:y}),a?null:(0,h.jsx)(l.I,{placeholder:D?"Job title*":"Job title",name:"title",required:D}),(0,h.jsx)(l.I,{placeholder:"Phone number",name:"phone",id:"phone",type:"tel"}),(0,h.jsx)(l.I,{type:"email",placeholder:"Work email*",name:"email",required:!0}),I]}),a||P?null:(0,h.jsx)(d.g,{placeholder:w,id:"00N7Q00000CkEBw",name:"00N7Q00000CkEBw",required:b,className:r.Z.textarea}),(0,h.jsx)(c.X,{label:a?(0,h.jsxs)(h.Fragment,{children:["I consent to allow Intertech Services AG and its affiliates to process my personal information provided in this form under the conditions"," ",(0,h.jsx)(o.p.Desktop,{children:(0,h.jsx)("br",{})}),"and for the purposes defined in the Privacy Policy and Customer Service Agreement, as well as to contact me for follow-up to this request."]}):(0,h.jsxs)(h.Fragment,{children:["I consent to allow Intertech Services AG (Werftestrasse 4, 6005 Luzern, Switzerland) to process my personal information provided in this form to contact me for follow-up to this request. If at any point you wish to withdraw your consent, please fill out the form"," ",(0,h.jsx)(u.r,{target:"_blank",href:"/unsubscribe/",children:"here"}),". For more information about what we do with personal data see our"," ",(0,h.jsx)(u.r,{target:"_blank",href:"/legal/confidential/",children:"privacy notice"}),"."]}),required:!0,className:r.Z.checkbox}),(0,h.jsx)(c.X,{label:a?(0,h.jsxs)(h.Fragment,{children:["I agree to receive advertising and other marketing emails from Intertech Services AG in accordance with the Privacy Policy"," ",(0,h.jsx)(o.p.Desktop,{children:(0,h.jsx)("br",{})})," ","and Customer Service Agreement, including newsletters, invitations to events, and promotional emails."]}):(0,h.jsxs)(h.Fragment,{children:["I agree to receive advertising and other marketing emails from Intertech Services AG (Werftestrasse 4, 6005 Luzern, Switzerland), including newsletters, invitations to events, and promotional emails. If you wish to unsubscribe you can do so by clicking on the unsubscribe link at the end of any email you have received from us to your email address. For more information about what we do with personal data see our"," ",(0,h.jsx)(u.r,{target:"_blank",href:"/legal/confidential/",children:"privacy notice"}),". For more information please click"," ",(0,h.jsx)(u.r,{target:"_blank",href:"/legal/e-newsletter/",children:"here"}),"."]}),id:"00N7Q00000CkDzy",name:"00N7Q00000CkDzy",className:r.Z.checkbox,setValue:"Allow",defaultValue:"Forbid"}),(0,h.jsx)(s.z,{view:"action",type:"submit",className:r.Z.button,children:S}),(0,h.jsx)("input",{id:"lead_source",name:"lead_source",type:"hidden",value:x})]})}},16318:function(e,t,n){n.r(t),n.d(t,{frontmatter:function(){return f},default:function(){return p}});var r=n(51512),a=n(86784),i=n(88555),s=n(34151),o=n(41427),l=n(46237),c=n(94501),d=n(4511),u=n(85893),f={slug:"tutorial-wsdm",title:"Tutorial at WSDM 2020",description:"In this tutorial, we present some key techniques for efficiently collecting labeled data, including aggregation, incremental relabeling, and dynamic pricing.",date:"2020-02-03T10:00:00.000Z",category:"conference",hosts:[{imageId:"CV2fFNwtYAmTaFnMcVTc",name:"Alexey Drutsa"},{imageId:"zGHkk2wbT4yuQVbfFaJO",name:"Valentina Fedorova"},{imageId:"VFeyhjwZLWoBnWODaDDa",name:"Olga Megorskaya"},{imageId:"oL04gaQuspW0NPZLsF8N",name:"Evfrosiniya Zerminova"},{imageId:"Rh4qN8gXmzlgyEkuGvz6",name:"Dmitry Ustalov"},{imageId:"YJ7-107LJUdfxbpwJB5E",name:"Daria Baidakova"}]};function p(e){var t=e.data;return(0,u.jsx)(s.r,{data:t,eventCategory:f.category,eventSlug:f.slug,eventDate:f.date,eventIsOnDemand:f.isOnDemand,children:(0,u.jsxs)(i.D,{toc:[{url:"#overview",title:"Overview"},{url:"#speakers",title:"Speakers"},{url:"#schedule",title:"Schedule"},{url:"#slides",title:"Slides"}],title:f.title,description:f.description,backLink:{text:"Back to Events",iconPosition:"left",href:"/events"},category:f.category,date:f.date,children:[(0,u.jsx)(l.D,{level:2,styleAs:4,id:"overview",children:"Overview"}),(0,u.jsx)("p",{children:"In this tutorial, we present portion of unique industry experience in efficient data labeling via crowdsourcing. The majority of ML projects require training data, and often this data can only be obtained by human labeling. Moreover, the more applications of AI appear, the more nontrivial tasks for collecting human labeled data arise. Production of such data on a large-scale requires construction of a technological pipeline, which includes solving issues related to quality control and smart distribution of tasks between performers."}),(0,u.jsx)("p",{children:"We introduce data labelling via public crowdsourcing marketplaces and present the key components of efficient label collection. This is followed by a practice session, where participants choose one real label collection task, experiment with selecting settings for the labeling process, and launch their label collection project on Toloka, one of the world's largest crowdsourcing marketplaces."}),(0,u.jsx)(r.w,{offset:56}),(0,u.jsx)(l.D,{level:2,styleAs:4,id:"speakers",children:"Speakers"}),(0,u.jsx)(o.B,{people:[{imgSrc:"CV2fFNwtYAmTaFnMcVTc",name:"Alexey Drutsa",company:"Toloka",info:"Head of Efficiency & Growth Division"},{imgSrc:"zGHkk2wbT4yuQVbfFaJO",name:"Valentina Fedorova",company:"Toloka",info:"Analyst"},{imgSrc:"VFeyhjwZLWoBnWODaDDa",name:"Olga Megorskaya",company:"Toloka",info:"CEO"},{imgSrc:"oL04gaQuspW0NPZLsF8N",name:"Evfrosiniya Zerminova",company:"Toloka",info:"Technical Product Manager"},{imgSrc:"Rh4qN8gXmzlgyEkuGvz6",name:"Dmitry Ustalov",company:"Toloka",info:"Head of Research"},{imgSrc:"YJ7-107LJUdfxbpwJB5E",name:"Daria Baidakova",company:"Toloka",info:"Director of Educational Programs"}]}),(0,u.jsx)(r.w,{offset:80}),(0,u.jsx)(l.D,{level:2,styleAs:4,id:"schedule",children:"Schedule"}),(0,u.jsx)(c.B,{content:[{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_introduction.pdf",target:"_blank",children:"Introduction"}),text:(0,u.jsxs)(u.Fragment,{children:["— The concept of crowdsourcing ",(0,u.jsx)("br",{}),"— Crowdsourcing task examples",(0,u.jsx)("br",{}),"— Crowdsourcing platforms",(0,u.jsx)("br",{}),"— Toloka crowdsourcing experience"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part1.pdf",target:"_blank",children:"Part I: Main components of data collection via crowdsourcing"}),text:(0,u.jsxs)(u.Fragment,{children:["— Decomposition for an effective pipeline",(0,u.jsx)("br",{}),"— Task instruction & interface: best practices",(0,u.jsx)("br",{}),"— Quality control techniques"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part2.pdf",target:"_blank",children:"Part II: Label collection projects (practical session)"}),text:(0,u.jsxs)(u.Fragment,{children:["— Dataset and required labels",(0,u.jsx)("br",{}),"— Discussion: how to collect labels?",(0,u.jsx)("br",{}),"— Data labeling pipeline for implementation"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part3.pdf",target:"_blank",children:"Part III: Introduction to Toloka for requesters"}),text:(0,u.jsxs)(u.Fragment,{children:["— Main types of instances",(0,u.jsx)("br",{}),"— Project: creation & configuration",(0,u.jsx)("br",{}),"— Pool: creation & configuration",(0,u.jsx)("br",{}),"— Tasks: uploading & golden set creation",(0,u.jsx)("br",{}),"— Statistics in flight and results downloading"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part4.pdf",target:"_blank",children:"Part IV: Setting up & running label collection projects"}),text:(0,u.jsxs)(u.Fragment,{children:["Participants:",(0,u.jsx)("br",{}),"— create",(0,u.jsx)("br",{}),"— configure",(0,u.jsx)("br",{}),"— run data labeling projects on real performers in real-time"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part5.pdf",target:"_blank",children:"Part V: Interface & quality control"}),text:(0,u.jsxs)(u.Fragment,{children:["— Detailed examination of quality control techniques",(0,u.jsx)("br",{}),"— Comprehensive overview of best practices for creating a functional interface"]})},{date:(0,u.jsxs)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part7.pdf",target:"_blank",children:["Part VII: Setting up and running label ",(0,u.jsx)("br",{})," collection projects cont. (practice session)"]}),text:(0,u.jsxs)(u.Fragment,{children:["Participants:",(0,u.jsx)("br",{}),"— create",(0,u.jsx)("br",{}),"— configure",(0,u.jsx)("br",{}),"— run data labeling projects on real performers in real-time"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part8.pdf",target:"_blank",children:"Part VIII: Theory on incremental relabeling and pricing"}),text:(0,u.jsxs)(u.Fragment,{children:["— Incremental relabeling to save money",(0,u.jsx)("br",{}),"— Performance-based pricing"]})},{date:(0,u.jsx)(d.r,{href:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part9.pdf",target:"_blank",children:"Part IX: Discussion of results from the projects and conclusions"}),text:(0,u.jsxs)(u.Fragment,{children:["— Results of your projects",(0,u.jsx)("br",{}),"— Ideas for further work and research",(0,u.jsx)("br",{}),"— References to literature and other tutorials"]})}]}),(0,u.jsx)(r.w,{offset:56}),(0,u.jsx)(l.D,{level:2,styleAs:4,id:"slides",children:"Slides"}),(0,u.jsx)(a._,{files:[{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_introduction.pdf",title:"Introduction"},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part1.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 1:"})," Main components of data collection"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part2.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 2:"})," Label collection projects to be done crowdsourcing"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part3.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 3:"})," Introduction to Toloka for requesters"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part4.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 4:"})," Setting up and running label collection projects"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part5.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 5:"})," Interface and quality control"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part6.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 6:"})," Theory on efficient aggregation projects"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part7.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 7:"})," Setting up and running label collection projects"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part8.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 8:"})," Theory on incremental relabelling and pricing"]})},{link:"https://tlk-infra-front.azureedge.net/portal-static/files/conference-presentations/wsdm/wsdm2020_part9.pdf",title:(0,u.jsxs)(u.Fragment,{children:[(0,u.jsx)("b",{children:"Part 9:"})," Results and conclusions"]})}]})]})})}}}]);