deploy: dd7b0d6

2015/10/02/the-wonderful-world-of-recommender-systems/index.html

@@ -7,7 +7,7 @@
 100vw" srcset="https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot_hu17008018622414803394.png 360w,
 https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot_hu16912368895272562788.png 480w,
 https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot_hu9102468640053471026.png 720w,
-https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot.png 750w," src=https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot.png alt="Hynt recommendation widget" loading=lazy></a></figure><p><a href=https://hynt.com target=_blank rel=noopener>Hynt</a> is a recommender-system-as-a-service for e-commerce whose development I led up until the middle of last year. The general idea is that merchants simply add a few lines of JavaScript to their shop pages and Hynt does the hard work of recommending relevant items from the store, while considering the user and page context. Going live with Hynt reaffirmed many well-known UI/UX lessons. Most notably:</p><ul><li><em>Above the fold is better than below.</em> Engagement with Hynt widgets that were visible without scrolling was higher than those that were lower on the page.</li><li><em>More recommendations are better than a few.</em> Hynt widgets are responsive, adapting to the size of the container they&rsquo;re placed in. Engagement was more likely when more recommendations were displayed, because users were more likely to find something they liked without scrolling through the widget.</li><li><em>Fast is better than slow.</em> If recommendations load faster, more people see them, which increases engagement. In Hynt&rsquo;s case speed was especially important because the widgets load asynchronously after the host page finishes loading.</li></ul><p>Another important UI/UX element is explanations. Displaying a plausible explanation next to a recommendation can do wonders, without making any changes to the underlying recommendation algorithms. The impact of explanations has been studied extensively by Nava Tintarev and Judith Masthoff. They have identified seven different aims of explanations, which are summarised in the following table (reproduced from their <a href=http://homepages.abdn.ac.uk/n.tintarev/pages/papers/TintarevMasthoffICDE07.pdf target=_blank rel=noopener>survey of explanations in recommender systems</a>).</p><table><thead><tr><th>Aim</th><th>Definition</th></tr></thead><tbody><tr><td>Transparency</td><td>Explain how the system works</td></tr><tr><td>Scrutability</td><td>Allow users to tell the system it is wrong</td></tr><tr><td>Trust</td><td>Increase user confidence in the system</td></tr><tr><td>Effectiveness</td><td>Help users make good decisions</td></tr><tr><td>Persuasiveness</td><td>Convince users to try or buy</td></tr><tr><td>Efficiency</td><td>Help users make decisions faster</td></tr><tr><td>Satisfaction</td><td>Increase ease of usability or enjoyment</td></tr></tbody></table><p>Explanations are ubiquitous in real-world recommender systems. For example, Amazon uses explanations like &ldquo;frequently bought together&rdquo;, and &ldquo;customers who bought this item also bought&rdquo;, while Netflix presents different lists of recommendations where each list is driven by a different reason. However, as the following Netflix example shows, it is worth making sure that the explanations you provide don&rsquo;t <a href=http://funnyjunk.com/Thanks+netflix/funny-pictures/5040772/ target=_blank rel=noopener>make you look stupid</a>.</p><figure><a href=amazon-frequently-bought-together.png target=_blank rel=noopener><img sizes="(min-width: 768px) 633px,
+https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot.png 750w," src=https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/hynt-screenshot.png alt="Hynt recommendation widget" loading=lazy></a></figure><p><a href=https://hynt.com target=_blank rel=noopener>Hynt</a> is a recommender-system-as-a-service for e-commerce whose development I led up until the middle of last year. The general idea is that merchants simply add a few lines of JavaScript to their shop pages and Hynt does the hard work of recommending relevant items from the store, while considering the user and page context. Going live with Hynt reaffirmed many well-known UI/UX lessons. Most notably:</p><ul><li><em>Above the fold is better than below.</em> Engagement with Hynt widgets that were visible without scrolling was higher than those that were lower on the page.</li><li><em>More recommendations are better than a few.</em> Hynt widgets are responsive, adapting to the size of the container they&rsquo;re placed in. Engagement was more likely when more recommendations were displayed, because users were more likely to find something they liked without scrolling through the widget.</li><li><em>Fast is better than slow.</em> If recommendations load faster, more people see them, which increases engagement. In Hynt&rsquo;s case speed was especially important because the widgets load asynchronously after the host page finishes loading.</li></ul><p>Another important UI/UX element is explanations. Displaying a plausible explanation next to a recommendation can do wonders, without making any changes to the underlying recommendation algorithms. The impact of explanations has been studied extensively by Nava Tintarev and Judith Masthoff. They have identified seven different aims of explanations, which are summarised in the following table (reproduced from their <a href=http://homepages.abdn.ac.uk/n.tintarev/pages/papers/TintarevMasthoffICDE07.pdf target=_blank rel=noopener>survey of explanations in recommender systems</a>).</p><table><thead><tr><th style=text-align:left>Aim</th><th style=text-align:left>Definition</th></tr></thead><tbody><tr><td style=text-align:left>Transparency</td><td style=text-align:left>Explain how the system works</td></tr><tr><td style=text-align:left>Scrutability</td><td style=text-align:left>Allow users to tell the system it is wrong</td></tr><tr><td style=text-align:left>Trust</td><td style=text-align:left>Increase user confidence in the system</td></tr><tr><td style=text-align:left>Effectiveness</td><td style=text-align:left>Help users make good decisions</td></tr><tr><td style=text-align:left>Persuasiveness</td><td style=text-align:left>Convince users to try or buy</td></tr><tr><td style=text-align:left>Efficiency</td><td style=text-align:left>Help users make decisions faster</td></tr><tr><td style=text-align:left>Satisfaction</td><td style=text-align:left>Increase ease of usability or enjoyment</td></tr></tbody></table><p>Explanations are ubiquitous in real-world recommender systems. For example, Amazon uses explanations like &ldquo;frequently bought together&rdquo;, and &ldquo;customers who bought this item also bought&rdquo;, while Netflix presents different lists of recommendations where each list is driven by a different reason. However, as the following Netflix example shows, it is worth making sure that the explanations you provide don&rsquo;t <a href=http://funnyjunk.com/Thanks+netflix/funny-pictures/5040772/ target=_blank rel=noopener>make you look stupid</a>.</p><figure><a href=amazon-frequently-bought-together.png target=_blank rel=noopener><img sizes="(min-width: 768px) 633px,
 100vw" srcset="https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/amazon-frequently-bought-together_hu15589678651873710813.png 360w,
 https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/amazon-frequently-bought-together_hu18021758773494046297.png 480w,
 https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/amazon-frequently-bought-together.png 633w," src=https://yanirseroussi.com/2015/10/02/the-wonderful-world-of-recommender-systems/amazon-frequently-bought-together.png alt="Amazon frequently bought together" loading=lazy></a></figure><figure><a href=netflix-because-you-watched.png target=_blank rel=noopener><img sizes="(min-width: 768px) 720px,

index.html

@@ -1,4 +1,4 @@
-<!doctype html><html lang=en dir=auto><head><meta name=generator content="Hugo 0.133.1"><meta charset=utf-8><meta http-equiv=X-UA-Compatible content="IE=edge"><meta name=viewport content="width=device-width,initial-scale=1,shrink-to-fit=no"><meta name=robots content="index, follow"><title>Yanir Seroussi | Data & AI for Startup Impact</title>
+<!doctype html><html lang=en dir=auto><head><meta name=generator content="Hugo 0.134.1"><meta charset=utf-8><meta http-equiv=X-UA-Compatible content="IE=edge"><meta name=viewport content="width=device-width,initial-scale=1,shrink-to-fit=no"><meta name=robots content="index, follow"><title>Yanir Seroussi | Data & AI for Startup Impact</title>
 <meta name=description content="Helping climate & nature tech startups ship data-intensive solutions (artificial intelligence, machine learning, data science, and advanced analytics).
 "><meta name=author content="Yanir Seroussi"><link rel=canonical href=https://yanirseroussi.com/><meta name=google-site-verification content="aWlue7NGcj4dQpjOKJF7YKiAvw3JuHnq6aFqX6VwWAU"><link crossorigin=anonymous href=/assets/css/stylesheet.6f5c97224af1f1714566202529b7d458386b85c4df858c71df30dd5c1c769363.css integrity="sha256-b1yXIkrx8XFFZiAlKbfUWDhrhcTfhYxx3zDdXBx2k2M=" rel="preload stylesheet" as=style><link rel=icon href=https://yanirseroussi.com/favicon.ico><link rel=icon type=image/png sizes=16x16 href=https://yanirseroussi.com/favicon-16x16.png><link rel=icon type=image/png sizes=32x32 href=https://yanirseroussi.com/favicon-32x32.png><link rel=apple-touch-icon href=https://yanirseroussi.com/apple-touch-icon.png><link rel=mask-icon href=https://yanirseroussi.com/safari-pinned-tab.svg><meta name=theme-color content="#2e2e33"><meta name=msapplication-TileColor content="#2e2e33"><link rel=alternate type=application/rss+xml href=https://yanirseroussi.com/index.xml><link rel=alternate hreflang=en href=https://yanirseroussi.com/><noscript><style>#theme-toggle,.top-link{display:none}</style><style>@media(prefers-color-scheme:dark){:root{--theme:rgb(29, 30, 32);--entry:rgb(46, 46, 51);--primary:rgb(218, 218, 219);--secondary:rgb(155, 156, 157);--tertiary:rgb(65, 66, 68);--content:rgb(196, 196, 197);--code-block-bg:rgb(46, 46, 51);--code-bg:rgb(55, 56, 62);--border:rgb(51, 51, 51)}.list{background:var(--theme)}.list:not(.dark)::-webkit-scrollbar-track{background:0 0}.list:not(.dark)::-webkit-scrollbar-thumb{border-color:var(--theme)}}</style></noscript><meta property="og:title" content="Yanir Seroussi | Data & AI for Startup Impact"><meta property="og:description" content="Helping climate & nature tech startups ship data-intensive solutions (artificial intelligence, machine learning, data science, and advanced analytics).
 "><meta property="og:type" content="website"><meta property="og:url" content="https://yanirseroussi.com/"><meta name=twitter:card content="summary"><meta name=twitter:title content="Yanir Seroussi | Data & AI for Startup Impact"><meta name=twitter:description content="Helping climate & nature tech startups ship data-intensive solutions (artificial intelligence, machine learning, data science, and advanced analytics).