{"id":4329,"date":"2021-04-17T01:10:26","date_gmt":"2021-04-17T01:10:26","guid":{"rendered":"https:\/\/dalelane.co.uk\/blog\/?p=4329"},"modified":"2021-04-17T01:12:28","modified_gmt":"2021-04-17T01:12:28","slug":"building-a-question-answering-game-in-scratch","status":"publish","type":"post","link":"https:\/\/dalelane.co.uk\/blog\/?p=4329","title":{"rendered":"Building a Question Answering game in Scratch"},"content":{"rendered":"<p>I added a new project worksheet to <a href=\"https:\/\/machinelearningforkids.co.uk\/worksheets\">Machine Learning for Kids<\/a> today.<\/p>\n<p>It has step-by-step instructions for how to make a <a href=\"https:\/\/github.com\/IBM\/taxinomitis-docs\/raw\/master\/project-worksheets\/pdf\/worksheet-quizshow.pdf\">quiz show game<\/a> in <a href=\"https:\/\/scratch.mit.edu\">Scratch<\/a> that uses a machine learning model to understand questions on any topic the student chooses, and find the answer in Wikipedia pages.<\/p>\n<p><img decoding=\"async\" src=\"\/\/dalelane.co.uk\/blog\/post-images\/210417-worksheet.png\" style=\"border: thin black solid\"\/><\/p>\n<p>It&#8217;s a fun little project, super simple to make, and works surprisingly well. It doesn&#8217;t get every question right, but it does a lot better than I expected.<\/p>\n<p><img decoding=\"async\" src=\"\/\/dalelane.co.uk\/blog\/post-images\/210417-scratch.png\" style=\"border: thin black solid\"\/><\/p>\n<p>I don&#8217;t normally write blog posts when I write new ML for Kids worksheets, but this one was a bit interesting.<\/p>\n<p><!--more-->The topic I chose to <a href=\"https:\/\/youtu.be\/R4OHKQpWYUM\">demo the project<\/a> was the Wimbledon tennis championships &#8211;  asking questions such as what colour balls they use at Wimbledon, when the retractable roof was first introduced, the spectator capacity of Centre Court, and so on.<\/p>\n<p><iframe loading=\"lazy\" width=\"450\" height=\"375\" src=\"https:\/\/www.youtube.com\/embed\/6Vo7xBfDdEU\" title=\"YouTube video player\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture\" allowfullscreen=\"\" style=\"border: thin black solid\"><\/iframe><br \/>\n<small><a href=\"https:\/\/youtu.be\/6Vo7xBfDdEU\">youtu.be\/6Vo7xBfDdEU<\/a><\/small><\/p>\n<p>The reason I wanted to mention it is because that was a bit of a callback.<\/p>\n<p>In June 2016, I <a href=\"https:\/\/dalelane.co.uk\/blog\/?p=3381\">shared a demo of something I was working on<\/a>, called <strong>Watson Retrieve &amp; Rank<\/strong>. In that demo, I showed how to train a machine learning system to find answers to questions in a corpus of documents. To do that I also used questions about the Wimbledon tennis championship, retrieving answers from a corpus of documents about Wimbledon.<\/p>\n<p><img decoding=\"async\" src=\"\/\/dalelane.co.uk\/blog\/post-images\/210417-randr.png\" style=\"border: thin black solid\"\/><\/p>\n<p><strong>These two demos give a nice snapshot into how far machine learning tech has progressed in the last five years.<\/strong><\/p>\n<p>The fact that you can now do this in JavaScript in a web browser without any dependencies on web services (the only Internet access in the project is where it retrieves the Wikipedia page &#8211; everything else is done on the student&#8217;s own computer running in the web browser) is just amazing.<\/p>\n<p><em>(To be clear, these two demos are not doing exactly the same job &#8211; in Retrieve and Rank, the QA system was finding an answer in a large corpus of documents, whereas in this student Scratch project, the QA system is finding an answer in a single page. But it&#8217;s similar. And still super impressive to me.)<\/em><\/p>\n<p><iframe loading=\"lazy\" width=\"450\" height=\"280\" src=\"https:\/\/www.youtube.com\/embed\/SveIksv7V9E\" title=\"YouTube video player\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture\" allowfullscreen=\"\" style=\"border: thin black solid\"><\/iframe><br \/>\n<small><a href=\"https:\/\/youtu.be\/6Vo7xBfDdEU\">youtu.be\/SveIksv7V9E<\/a> (June 2016)<\/small><\/p>\n<p>And that wasn&#8217;t even the first time I worked on a question answering system that could answer questions about Wimbledon. In 2015, I worked on <a href=\"https:\/\/newsroom.ibm.com\/2015-06-18-Wimbledon-and-IBM-Push-Digital-Boundaries-to-Enhance-Fan-and-Player-Engagement,1\">a project for Wimbledon using something called <strong>Watson Engagement Advisor<\/strong><\/a>:<\/p>\n<blockquote><p>&#8230; Watson Engagement Advisor provides related insights and historical context. Wimbledon staff will be able to pose questions in natural language as if they had the world&#8217;s best tennis expert on-hand, and share these insights with fans via social media and the Wimbledon digital platforms &#8230; [This] will bring unprecedented analysis &#8230; For example, Wimbledon staff will be able to quickly surface information and insights about interesting or record-breaking player and match statistics<\/p><\/blockquote>\n<p>(I don&#8217;t have any screenshots of that project to hand, but you can see some in <a href=\"https:\/\/www.linkedin.com\/pulse\/wimbledon-2015-amazing-ibm-technology-behind-tennis-nathaniel\/\">this blog post<\/a>.)<\/p>\n<p>Essentially, I&#8217;ve spent a <strong>lot<\/strong> of time making question answering systems about the Wimbledon championships. So it was a fun bit of question answering nostalgia for me to revisit that again, this time to enable children to do it!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>I added a new project worksheet to Machine Learning for Kids today. It has step-by-step instructions for how to make a quiz show game in Scratch that uses a machine learning model to understand questions on any topic the student chooses, and find the answer in Wikipedia pages. It&#8217;s a fun little project, super simple [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":4336,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[578,536,505],"class_list":["post-4329","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tech","tag-ibmwatson","tag-scratch","tag-watson"],"_links":{"self":[{"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=\/wp\/v2\/posts\/4329","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4329"}],"version-history":[{"count":0,"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=\/wp\/v2\/posts\/4329\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=\/wp\/v2\/media\/4336"}],"wp:attachment":[{"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4329"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4329"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dalelane.co.uk\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4329"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}