{"id":34,"date":"2020-03-13T09:12:45","date_gmt":"2020-03-13T09:12:45","guid":{"rendered":"http:\/\/yom-tov.info\/blog\/?p=34"},"modified":"2020-03-13T09:12:45","modified_gmt":"2020-03-13T09:12:45","slug":"this-ad-may-save-your-life","status":"publish","type":"post","link":"https:\/\/yom-tov.info\/blog\/2020\/03\/13\/this-ad-may-save-your-life\/","title":{"rendered":"This ad may save your life"},"content":{"rendered":"\n

Internet advertising systems know a lot about us. If you\nwant to know how much, head to Google’s web page on ad personalization (https:\/\/adssettings.google.com\/u\/0\/authenticated<\/a>).\nOn a recent visit I found out that Google knows of my upcoming travel plans to\nItaly and Texas, a few of my hobbies, and several academic topics that I\u2019m\nlearning more about. Unsurprisingly, it wasn’t correct on everything (I\u2019m not\ninto extreme sports), but it knew a lot more than I would have imagined before\nI visited their web page.<\/p>\n\n\n\n

Over the past few years our and other research groups have shown that interactions with search\nengines can be used to screen people for a variety of serious medical\nconditions, both mental and physical. These include depression, eating\ndisorders, Parkinson’s disease, several types of solid tumor cancer, and\ndiabetes. However, informing people about these inferences is challenging both\ntechnically and ethically.<\/p>\n\n\n\n

This week we published a paper (https:\/\/dl.acm.org\/doi\/10.1145\/3373720<\/a>)\nshowing how to leverage the information that Internet advertisers have about us,\nto screen for 3 types of cancer. Our results suggest that it\u2019s indeed possible\nto screen people for their likelihood of suffering from cancer before they are\ndiagnosed by a doctor. <\/p>\n\n\n\n

Here’s how it works: when an advertiser uses Bing or Google\nto advertise, they select keywords such that when a user searches for these\nkeywords their ads are shown. A\nmore sophisticated form\nof advertising happens when, in addition, advertisers tell Google or Bing whenever a user who saw the ads buys the product they\nwere trying to sell. When advertisers do this, the advertising systems learn to\npredict who, among all people use the keywords, are likely to buy a product (technically\nthis is known as conversion optimization). This learning is based on the\ninformation that advertising systems have about users, including their\ninterests, locations and demographics.<\/p>\n\n\n\n

What we did was to leverage this capability and use the\nadvertising system to screen people for cancer. We achieved this by showing an ad to people who searched\nfor information on self-diagnosis of lung, breast or colon cancers. The ad suggested help in understanding the\nseverity of the symptoms that people were experiencing. People who chose to click\nthe ads were directed to a website where, after explaining the experimental\nnature of the system and asking for their consent, they were given a clinical\nquestionnaire about their demographics and symptoms. People who answered the questions were\ngiven one of two indications: either that they should urgently seek medical\nattention, because their symptoms were deemed serious, or that it was likely\nthat their symptoms weren\u2019t indicative of cancer but medical advice should be\nsought, though not urgently. <\/p>\n\n\n\n

When the questionnaire indicated that a person was likely\nsuffering from cancer, we informed the advertising system that the person \u201cbought\u201d\nour \u201cproduct\u201d. Within 3 weeks, the advertising systems learned to focus on\nthose people who probably have cancer, such that approximately 1 in 10 people who\ncompleted the questionnaires were likely suffering from it, up from the\nbaseline rate of under 1%. This rate was similar for all three types of cancer.<\/p>\n\n\n\n

The use of ads offers a method for interacting with people\nwho might be suffering from as-of-yet\nundiagnosed cancer. By providing ads with an offer of help and\nempowering people to select whether or not they wished to receive this help, we\novercome many of the ethical challenges associated with unsolicited diagnosis.\nOur use of the sophisticated capabilities and knowledge about users that\nadvertising systems have, allows us to identify people with serious disease,\nwithout having to have access to sensitive individual-level search data.<\/p>\n\n\n\n

Interestingly the people who use the system most came from\ncountries with high Internet use and lower life span. The latter is a known proxy for the\nquality of the health system. <\/p>\n\n\n\n

Many health organizations use internet advertising for awareness\ncampaigns and for campaigns designed to encourage healthier behaviors. Our results lead us to\nsuggest that health systems should leverage the information that advertising\nsystems collect about people in order to improve population level screening\nprograms. <\/p>\n\n\n\n

Jeffrey Hammerbacher (at the time at Facebook) once commented\nthat \u201cThe best minds of my generation are thinking about how to make people\nclick ads. That sucks.\u201d Let’s make use of the products of those great minds to\nimprove outcomes for people with serious disease.<\/p>\n","protected":false},"excerpt":{"rendered":"

Internet advertising systems know a lot about us. If you want to know how much, head to Google’s web page on ad personalization (https:\/\/adssettings.google.com\/u\/0\/authenticated). On a recent visit I found out that Google knows of my upcoming travel plans to Italy and Texas, a few of my hobbies, and several academic topics that I\u2019m learning … <\/p>\n

Continue reading “This ad may save your life”<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-34","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/posts\/34","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/comments?post=34"}],"version-history":[{"count":0,"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/posts\/34\/revisions"}],"wp:attachment":[{"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/media?parent=34"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/categories?post=34"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/yom-tov.info\/blog\/wp-json\/wp\/v2\/tags?post=34"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}