{"id":677715,"date":"2019-07-01T09:00:09","date_gmt":"2019-07-01T16:00:09","guid":{"rendered":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/?post_type=msr-blog-post&#038;p=677715"},"modified":"2020-07-23T11:15:48","modified_gmt":"2020-07-23T18:15:48","slug":"user-research-makes-your-ai-smarter","status":"publish","type":"msr-blog-post","link":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/articles\/user-research-makes-your-ai-smarter\/","title":{"rendered":"User research makes your AI smarter"},"content":{"rendered":"<p>By <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/www.linkedin.com\/in\/pennycollisson\/\">Penny Marsh Collisson<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> and <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/www.linkedin.com\/in\/ghardiman\/\">Gwyneth Hardiman<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<div id=\"attachment_677727\" style=\"width: 634px\" class=\"wp-caption alignnone\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-677727\" class=\"wp-image-677727 size-full\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2020\/07\/Smarter-AI_header-image.png\" alt=\"Cartoon image with young person raising their right hand. Background is graphic image of shapes and design of various colors.\" width=\"624\" height=\"609\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2020\/07\/Smarter-AI_header-image.png 624w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2020\/07\/Smarter-AI_header-image-300x293.png 300w\" sizes=\"auto, (max-width: 624px) 100vw, 624px\" \/><p id=\"caption-attachment-677727\" class=\"wp-caption-text\">Note: This article was originally published on July 1, 2019, on Medium<\/p><\/div>\n<h3 id=\"6574\" class=\"hu hc bj bi ei hv hw hx hy hz ia ib ic id ie if ig ih ii ij ik bn\">Some things we\u2019re learning about doing UX research on AI at Microsoft<\/h3>\n<div class=\"il\">\n<div class=\"n fr im in io\">\n<div class=\"o n\">\n<p id=\"f240\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">As AI grows more prevalent, it\u2019s changing what people expect from technology and how they interact with it. That means that every UX-er and customer-obsessed product person needs to consider not only how to\u00a0<a class=\"bx dm lh li lj lk\" href=\"https:\/\/medium.com\/microsoft-design\/ai-guidelines-in-the-creative-process-807b6d31cda2?source=friends_link&sk=a942f71c0ca9ccc5129687cdd1042fb7\" target=\"_blank\" rel=\"noopener noreferrer\">create effective AI experiences<\/a>, but also how to collect customer feedback along the way. Easy\u2026right?<\/p>\n<p id=\"32dc\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">The good news is that many traditional research tools can help gauge customer reactions to AI. Ethnographies, focus groups, prototype research, customer surveys, and logs are all still relevant.\u00a0However, AI systems differ from classical systems in that they\u2019re context aware, personal, and able to learn over time. They also have limited accuracy and unique failure modes.\u00a0These things introduce new challenges and opportunities when researching the UX of AI.<\/p>\n<p id=\"5828\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">Today we\u2019ll share practical tips for researching the UX of AI that we\u2019ve learned along the way at Microsoft.<\/p>\n<h3 id=\"5f81\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Diversify your recruit<\/strong><\/h3>\n<div id=\"attachment_677733\" style=\"width: 410px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-677733\" class=\"wp-image-677733 size-full\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-2.png\" alt=\"Foreground: Pencil drawings of seven people, both men and women. Background a large green circle.\" width=\"400\" height=\"418\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-2.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-2-287x300.png 287w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><p id=\"caption-attachment-677733\" class=\"wp-caption-text\">Illustrations by Michaelvincent Santos<\/p><\/div>\n<p id=\"6adf\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">As UX-ers, it\u2019s our responsibility to ensure that the experiences we deliver embrace diversity and respect multiple contexts and capabilities. That\u2019s especially important with AI. If your AI UX is only usable for a subset of users, potentially harmful bias will creep into your AI models.\u00a0An arbitrary sample of participants, or even a split on basic demographics like gender and age, will not be enough to ensure your AI is inclusive.<\/p>\n<p id=\"4632\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\"><em>Even during early feedback stages, recruit for a wide array of characteristics such as these:<\/em><\/p>\n<ul class=\"\">\n<li id=\"f2f6\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Attitudes toward AI and privacy<\/li>\n<li id=\"3168\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Profiles of tech adoption<\/li>\n<li id=\"3808\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Levels of tech self-efficacy<\/li>\n<li id=\"d9b4\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Geographies<\/li>\n<li id=\"4077\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Social contexts and norms<\/li>\n<li id=\"a67a\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Physical, cognitive, or emotional abilities<\/li>\n<\/ul>\n<h3 id=\"2950\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Fake it till you make it with Wizard of Oz techniques<\/strong><\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-677739 size-full\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-3.png\" alt=\"Foreground: The back of a person looking at a large orange image. Background: A large orange image with a face traced in the color\" width=\"400\" height=\"523\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-3.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-3-229x300.png 229w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/p>\n<p>During early prototyping stages, it can be hard to get a good read on how your AI is going to work for people. Your prototype might be missing key functionality or interactivity that will impact how participants respond to the AI experience.<strong class=\"kc ll\">\u00a0<\/strong><a class=\"bx dm lh li lj lk\" href=\"https:\/\/en.wikipedia.org\/wiki\/Wizard_of_Oz_experiment\" target=\"_blank\" rel=\"noopener nofollow noreferrer\"><strong class=\"kc ll\">Wizard of Oz studies<\/strong><\/a> have participants interact with what they believe to be an AI system, while a human \u201cbehind the curtain\u201d simulates the behavior that the AI system would demonstrate.<\/p>\n<p>For example, a participant might think the system is providing recommendations based on her previous selections when a person in another room is actually providing them. When people can earnestly engage with what they\u00a0<em class=\"kp\">perceive<\/em>\u00a0to be an AI, they will form more complete mental models, while interacting with the experience in more natural ways.<\/p>\n<h3 id=\"e614\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Integrate people\u2019s real stuff into your AI prototype<\/strong><\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-677742 size-full alignright\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-4.png\" alt=\"A cartoon image of a person juggles several paper like documents that are connected by red lines\" width=\"400\" height=\"525\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-4.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-4-229x300.png 229w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/p>\n<p>If study participants see generic content, their reactions may mislead you.\u00a0People respond differently when the experience includes their real, personalized content such as photos, contacts, and documents.<\/p>\n<p>Imagine how you feel about a program that automatically detects faces in photographs. Now, imagine seeing the faces of your loved ones identified by the system. Your reaction may be very different when you see people you know. You\u2019ll need to spend extra time pre-populating\u00a0<a class=\"bx dm lh li lj lk\" href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/group\/customer-insights-research\/articles\/prototyping-empathy-six-tips\/\" target=\"_blank\" rel=\"noopener noreferrer\">your prototype<\/a>\u00a0with people\u2019s \u201creal\u201d content, but it will be worth the effort.<\/p>\n<h3><\/h3>\n<h3 id=\"e317\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Reference a person instead of an AI<\/strong><\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-677745 size-full alignleft\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-5.png\" alt=\"Background: Green and yellow silhouette of a person's facial profile. Foreground: The same person stands within their silhouette \" width=\"400\" height=\"425\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-5.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-5-282x300.png 282w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/p>\n<p id=\"4699\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">AI has a lot of hype and folklore around it. For that reason, referencing AI can cue participants to make certain assumptions \u2014 both good and bad \u2014 about their experience. For example, participants might key in on highly publicized stories of bias or failure in AI systems. Or they could assume AI is more capable and perfect than it will ever be.\u00a0Getting participants to think about how a\u00a0<a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/medium.com\/@jdlovejoy\/human-centered-ai-cheat-sheet-1da130ba1bab\" target=\"_blank\" rel=\"noopener noreferrer\">human could help them<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0can be a good way to glean insight about where AI can be useful.<\/p>\n<p id=\"5eb1\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\"><em>Here are some alternatives to talking about AI:<\/em><\/p>\n<ul class=\"\">\n<li id=\"6178\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Invite participants to share how they currently enlist other people to achieve their goals.<\/li>\n<li id=\"d5b5\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Ask participants how they would want a human expert to behave.<\/li>\n<\/ul>\n<h3 id=\"5a19\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Understand the impact of your AI getting it wrong<\/strong><\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-677757 size-full\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-6.png\" alt=\"A drawing of a person stands in a purple circle, streaks of blue colors surround them\" width=\"400\" height=\"529\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-6.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-6-227x300.png 227w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/p>\n<p id=\"0250\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">AI isn\u2019t perfect. It\u2019s probabilistic, fallible, and will make mistakes. Especially early in the design cycle, it can be easy to create perfect prototypes and get an overly optimistic response to your UX. While planning evaluations, build in realistic quirks or pitfalls to bridge the gulf between the shiny concept and realistic product execution.\u00a0Once you understand how your AI\u2019s failure modes impact people, you can design to mitigate their impact.<\/p>\n<p id=\"b4d9\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\"><em class=\"kp\">Here are a few methods to consider:<\/em><\/p>\n<ul class=\"\">\n<li id=\"ac4b\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Intentionally introduce things into your prototype that are likely to be \u201cwrong.\u201d<\/li>\n<li id=\"5c1d\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Ensure that system interactions in your Wizard of Oz studies\u00a0<a class=\"bx dm lh li lj lk\" href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/02\/Sketching-NLP.pdf\" target=\"_blank\" rel=\"noopener nofollow noreferrer\">include different kinds of errors<\/a>.<\/li>\n<li id=\"13bc\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Take participants down different paths: things are right, a little right, a little wrong, totally wrong.<\/li>\n<li id=\"2157\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Invite conversation about where failures would be most impactful to their experience.<\/li>\n<\/ul>\n<h3 id=\"4fe9\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Dive into mental models<\/strong><\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-677760 size-full alignleft\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-7.png\" alt=\"A drawing of a woman snorkeling in a large red and orange circle\" width=\"400\" height=\"512\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-7.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-7-234x300.png 234w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/p>\n<p id=\"973d\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\">People don\u2019t need to understand the nuts and bolts behind the technology powering AI to have a positive experience with it.But they need a mental model of the system \u2014 what it delivers, when, and why \u2014 to have realistic expectations of the system\u2019s capabilities.\u00a0It can be easy to assume that people correctly understand how your AI works, when frequently their understanding is wrong (even if they\u2019re confident about it!). Once we locate the gaps in people\u2019s mental models, we\u2019re better equipped to shore them up with our designs.<\/p>\n<p id=\"2248\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn gv cx\" data-selectable-paragraph=\"\"><em class=\"kp\">To understand how participants envision your AI system, try this:<\/em><\/p>\n<ul class=\"\">\n<li id=\"2841\" class=\"ka kb bj kc b hv kd ke hy kf kg kh ki id kj kk ig kl km ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Ask participants to write down the \u201crules\u201d for how the system works. For example, give them a result and ask them to explain why and how the system produced it.<\/li>\n<li id=\"ae4c\" class=\"ka kb bj kc b hv mj ke hy mk kg kh ml id kj mm ig kl mn ij kn mg mh mi cx\" data-selectable-paragraph=\"\">Have participants imagine that a human gave them a specific result.\u00a0Ask what it is about the data, or their interactions, that would have caused the human to give them that result.<\/li>\n<\/ul>\n<h3 id=\"06b1\" class=\"lm ln bj bi fs lo lp lq lr ls lt lu lv lw lx ly lz ma mb mc md cx\"><strong>Highlight exceptions as well as trends<\/strong><\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-677763 size-full aligncenter\" src=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-8.png\" alt=\"Sketch of a person reaching up to an orange block. Blue background.\" width=\"400\" height=\"405\" srcset=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-8.png 400w, https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-content\/uploads\/2019\/07\/Smarter-AI_image-8-296x300.png 296w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/p>\n<p>People will have different experiences with an AI depending on their context, the content they bring in, and the way they interact with the system.\u00a0There are challenges with extracting qualitative insights around AI systems based on what\u00a0<em class=\"kp\">most<\/em>\u00a0people do, or how they react, when every person\u2019s experience is so personal.\u00a0As you roll up results, pay close attention to outliers. Understand why participants had the unique experience they did within your sample. This is particularly important when evaluating the experience across a diverse audience.<\/p>\n<p><strong>What are some things you&#8217;ve learned about doing UX research on AI? Tweet us your thoughts <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"http:\/\/www.x.com\/MicrosoftRI\">@MicrosoftRI<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> or like us <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"http:\/\/www.facebook.com\/MicrosoftRI\">on Facebook<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> and join the conversation.<\/strong><\/p>\n<p id=\"9304\" class=\"ka kb bj kc b hv mu ke hy mv kg kh mw id kj mx ig kl my ij kn gv cx\" data-selectable-paragraph=\"\"><em>Gwenyth Hardiman is a senior design researcher. Penny Collisson is a user research manager working on AI and Assistant in Office and Windows.\u00a0<\/em><\/p>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>As artificial intelligence grows more prevalent in our products, it&#8217;s changing what our customers expect from their technology and how they interact with it. For user researches that means they need to know not only how to create effective experiences, but also be thinking of the customer. Here are a few tips to help you get started.<\/p>\n","protected":false},"author":39057,"featured_media":677790,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-content-parent":616842,"msr_hide_image_in_river":0,"footnotes":""},"research-area":[],"msr-locale":[268875],"msr-post-option":[],"class_list":["post-677715","msr-blog-post","type-msr-blog-post","status-publish","has-post-thumbnail","hentry","msr-locale-en_us"],"msr_assoc_parent":{"id":616842,"type":"group"},"_links":{"self":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/677715","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post"}],"about":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-blog-post"}],"author":[{"embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/users\/39057"}],"version-history":[{"count":12,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/677715\/revisions"}],"predecessor-version":[{"id":677814,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/677715\/revisions\/677814"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/media\/677790"}],"wp:attachment":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/media?parent=677715"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=677715"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=677715"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=677715"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}