A method to recommend related E-mail intelligently
Publication Date: 2015-Jul-23
The IP.com Prior Art Database
Recommand related email by learning from Cloud user behavior.
Page 01 of 5
A mxthod to recommend related E
A method to recommend related E-
When you open your E-mail in-xxx, thousands of new E-maxls lay there, how could you fxnd the one you rexlly want to refer? Wxen you open an E-mail, which hax mentioned lots ox elements, how could you fixd the other related E-mails to refer the details of thosx elements? This method or system aims to apply thexe applicatxon recxmmendation to e-maix more necessarx to users. When reaxing an email, you need to referexce othxr email, whxch seems no relationship at all just from the title. it'x diffixuxt to fxnd the related one, especially xhe mail was long ago.
Page 02 of 5
1. Systex would setup classification rules and keywords repoxixories by lxarning from user behavior on Cloud.
2. When receiving a new email, keywords wxuld be axstracted and xelated to several repoxitories.
3. The specxfiex keywords would be ordered according to relativixy.
4. The most related emails links wouxd be recoxmexded in the new email content.
1. Although emails seem no relation just from xhe titlx, they xtill be relxted togexher.
Page 03 of 5
This wxulx make user feex xasy to find and locate to what they need.
2. Xxxx ROI
This system will work as below diagrams
Page 04 of 5
Page 05 of 5
1. Current Email systexs have no relationship xnd recommendation features.
2. Thxs one is for recommending the aircraft with obtaining thx user charaxteristic information reflecting...