Overview

We advertised to people in Israel who made COVID-19-related symptom queries through the Google Ads platform. People who clicked on the ads were referred to the Microsoft Healthcare Bot. The interaction with the bot helped people understand the severity of their symptoms vis-à-vis their demographics and underlying medical conditions. The COVID19 instance of the Bot was created and managed by Israel’s second-largest hospital.

The advertising system was set to maximize the number of conversions. People who “converted” were defined as those people whose responses to the Healthcare Bot indicated that they were in need of urgent medical care.

Note that the algorithm used by the Healthcare Bot was independent of the advertising system. Therefore, the only way that conversion rates could be increased by the advertising system is by identifying people who were at higher likelihood for hospitalization according to the information that the ads system had about its users.

Analysis of user interactions with the ads and the Healthcare Bot at both the country and city levels helped predict COVID-19 case numbers and hospitalizations.

Search advertising campaign

Search advertising campaign1s consist of three main elements: keywords that, if used by a searcher, will trigger the appearance of an ad; the text of the ads; and the landing page, which is the page that a user will be referred to by clicking on an ad.

Ads were shown on the Google search engine to people in Israel. Google Adwords specifies that a user’s physical location is determined by IP address or by device location (see https://support.google.com/google-ads/answer/2453995). The ads were shown in response to queries containing one of the following terms (in Hebrew): dry cough, fever, stuffy nose, difficulty breathing, diarrhea, tiredness, headache, nausea, throat pain, or vomiting. The list of symptoms was based on the United Kingdom National Health Service’s first few hundred (FF100) survey on COVID-1919.

The ads comprised a short (one sentence) title and a 1–3 sentence body. The ads are shown in Table 1. In practice, Google forbade ads which mentioned COVID-19, and therefore two ads were not shown by the ad system.

The campaign was run in Hebrew.

Once the elements of the campaign are decided upon (e.g., the text of the ads is prepared) setting up the campaign requires 1–2 h.

Google AdWords provides aggregated information on the results of the ad campaign, including measures of the campaign (number of impressions, clicks, and conversions) over time, stratified by city (e.g., Haifa, Jerusalem, etc.), gender, and age group (18–24, 25–34, 35–44, 45–54, 55–64, and 65 and over). We denote the number of clicks out of all impressions as the clickthrough rate (CTR). The rate of conversions out of all clicks is denoted by the conversion rate (ConvR). Aggregated information can be downloaded at any time, and its processing, to predict, e.g., future hospitalization rates, can occur in near real-time.

Table 1 Advertisement text.

Microsoft health bot

People who clicked on the campaign ads were referred to an instance of the Microsoft Healthcare Bot (https://www.microsoft.com/en-us/research/project/health-bot/) which was installed by the Tel Aviv Sourasky Medical Center, Israel’s second largest hospital. The bot was configured by medical professionals to ask people about their demographics, underlying health conditions, and the symptoms they were experiencing. At the end of the interaction, the bot provided people with one of three outcomes: stay at home, call the family physician, or go to an emergency room. The latter was used as a feedback signal to the conversion optimization mechanism.

Ground truth case data

We retrospectively compared data from the advertising campaign to the Israeli Ministry of Health’s COVID-19 data repository (https://data.gov.il/dataset/covid-19). Specifically, our data was compared to the COVID-19 by area dataset, which provides information on the number of new COVID-19 cases per 100,000 people in each city on each day and the number of hospitalizations at these temporal and spatial resolutions. We refer to these as the case rate and hospitalization rate, respectively.

We note that case counts and hospitalizations are known to be noisy proxies for disease incidence, affected by policy, resources, and test accuracy. However, at the time of writing, these are the most comprehensive publicly available city-level data on COVID-19 in Israel.

Demographic data

We collected city-level demographic data from two official sources: Total population and the percentage of Arab citizens in a city from the Israeli Statistic Bureau20 and the median monthly income (2011), percentage of people earning less than the minimum wage and Gini index from the Israel National Insurance Institute21. These were used to evaluate differences in the demographics of people who utilized the campaign.

Data analysis

The correlation between the advertising-system measures and the pandemic indicators, taking into account the share of keywords for which the campaign ads were shown, was estimated through a linear model operating at daily granularity. The independent parameters of the model were either CTR or ConvR, expenditure per day, and the share of search impressions that day. The dependent parameters of the model were the case rates and hospitalization rates. Since it is hypothesized that campaign data may predict future indicators of the pandemic, the model is estimated at different lags between the data of the advertising system and those of the pandemic indicators, where a negative lag of k days means that data from the advertising system is correlated with pandemic data taken k days later. Models were estimated for lags ranging from − 14 to + 14 days.

By admin

https://muehlenbar.de/wp-includes/slot-online/

https://the-chef.co/sbobet/

Slot Gacor Terbaru

Link Slot Gacor

Sbobet

Situs Slot Gacor

Slot Gacor 2022

Login Sbobet

Daftar Sbobet

https://www.dcosmeticclinics.com.au/wp-includes/sbobet/

https://thetastesoflife.com/wp-includes/sbobet/

https://www.townshipofsugargrove.com/wp-includes/slot-gacor/

https://texasmamaboutique.com/wp-includes/slot-gacor/

https://bizu-me.com/wp-includes/slot-gacor/

https://tiketa.co.za/wp-includes/slot-gacor/

situs slot gacor

slot gacor gampang menang

slot gacor terbaru

daftar sbobet

slot bonus newmember

link sbobet

sbobet login