What are we collecting? 

The purpose of this study is to collect Time-Sensitive emails (TSEs) and Transactional emails to help improve anti-spam and email filtering technology.   

Time-sensitive emails are messages that contain information or requests that are relevant only for a specific period. Incorrect filtering of these emails can have negative consequences. Their value is greatly improved if these emails go to recipients’ inboxes and are not flagged as spam.

 
Transactional emails are emails sent to one individual at a time that specify or give further information about recent transactions, financial statements, shipment deliveries, receipts of purchases and travel bookings. 

Examples of TSEs are updates to a flight, outages for utility services, meeting invites or updates, reminders for in-person appointments, etc. 

What IS and what is NOT considered a TSE?  

Emails are considered time-sensitive whenever their delivery at a specific time is crucial, and not receiving or reading these in a timely manner could bring consequences to the recipient. These are generally (but not always) associated with payment due dates, appointments, or other calls to action.  

Emails containing advertisement, promotional emails, or emails that are sent in masse and are not specific to the recipient are NOT considered TSEs. Examples of non-time sensitive emails are discounts or promotions, flash sales, job applications or job alerts, newsletters, order confirmations, news, e-tickets or boarding passes, proof of payment, receipts, and payment confirmations.  

One quick way to confirm if an email is advertisement or promotional is to ask whether the email was sent because the recipient was part of a list (and is therefore generally an unsubscribe option in the email). Other potentially red flags that would indicate a promotional email are subjects including emojis, questions, mentions to saving money, sales, discounts, phrases like “you have been selected,” “congratulations,” or “don’t miss this opportunity.”  

Because the purpose of this project is to improve email filtering, the clear purpose of the email must be included in its subject. An email may contain time-sensitive information (e.g., a due date) but if its subject does not clearly define such information (e.g., “please read this email”), it is not considered a TSE for the purposes of this study.  

What do I need to check in my emails? 

For your emails to be eligible for the Project, they need to meet all the following criteria:  

• They must be in English. 

• Subject should have at least 4 words (excluding dates, numbers, tracking codes, or any other similar information, e.g., "Your flight AA123 time" only contains three words since the flight number does not count towards the wordcount). 

• Sender Email should have corporate domains (i.e., emails coming from addresses with @gmail, @yahoo, @aol, @outlook, @icloud or similar domains are not acceptable). 

• They must have Sender Name. 

• They must belong to one of the categories detailed in the Categories Section. 

• They must be in either .eml or .msg formats. 

• They must not be with subjects that contain Re: or Fw: or variations. 

• They must not be with subjects that contain emojis. 

• They must be no older than January 1st, 2021. 

• They must not be from a domain or with a subject included in the “Completed List” which you can check in the Domain and Subject List section. 

• The subject must provide enough information to deem the email TSE. General or generic subjects, as well as subject not providing enough information, will be rejected. 

• They must not be duplicate emails. Emails with identical or nearly identical subjects (e.g., utility notice from the same company for different months) are not accepted. 

• Please note that your emails should be varied: emails with similar subjects and only minimal changes (e.g., if you upload an email with the subject “Your bank statement for April”, and another one with the subject “Your bank statement for March”) will be considered duplicates and rejected. 

• They must be sent by a business or organization. 

• They must not be of a promotional nature (promote or advertise specific products or services). Some of the red flags are to check for promotional emails are: “make money”, “last days”, “savings” or “save”, “sale”, “don’t miss”, “awaits you”, “be part of”, “you’re selected”, “discount”, “congratulations”, “hurry (up)”, “(for) free”, “gift(s)”) or an ad (e.g., job ads, news, etc.). 

• They must not have subjects that are too general or don't provide enough information to determine whether email is TSE/Transactional or not (e.g., “new policy for lost baggage” or “please read this email”). 

• They must not have subjects with only changes in numbers from the same sender  

• Your booking is completed. Confirmation#: 1234 sender: confirm@booking.com    

• Your booking is completed. Confirmation#: 1235 sender: confirm@booking.com  

• They must not have subjects with only changes in dates from the same sender  

• Your appointment confirmation on Monday at 12:00 PM sender: dental@medicalgroup.com  

• Your appointment confirmation on Friday at 1:00 PM sender: dental@medicalgroup.com  

• They must belong to one of the categories we are collecting (e.g., the following categories are not accepted: OTPs, password recovery, delivery or package statuses). 

For this project we’re looking for a specific set of categories of emails, which you can see in the Categories section. These also include some examples of what is expected from those categories and what types of emails may be mistaken as part of those categories but are not actually in them.   

 
Please note that there are emails that are considered TSEs but we are ONLY looking for the ones shown in the Categories section.