Name-based classification of electronic account users
View Patent ↗Communication accounts may contain information such as account holder names, contact lists, and communication logs. Such information may be processed for generating features that may be used as corpus for a machine learning algorithm for developing classifiers of names. Specifically, names and contact names of the accounts may be arranged in to a document according to the manner in which account holders communicate with the contacts. The document may be used for generating word embedding of the names. Names prelabeled with ethnicity together with their word embedding may be used as training data for developing ethnicity classifiers based on machine learning algorithms.
1. A computer-implemented method comprising:
extracting a first account holder name from a first electronic communication account;
extracting a second account holder name from a second electronic communication account;
extracting first contact names from a first contact list of the first electronic communication account;
extracting second contact names from a second contact list of the second electronic communication account;
sorting the first contact names in the first contact list to generate first sorted contact names;
sorting the second contact names in the second contact list to generate second sorted contact names;
creating, from the first sorted contact names and the second sorted contact names, a document comprising a plurality of words representing the first account holder name, the second account holder name, the first contact names and the second contact names;
generating a word embedding of a defined number of dimensions from the document for one or more words of the plurality of words;
identifying a first country corresponding to a first registration country for the first electronic communication account;
identifying a second country corresponding to a second registration country for the second electronic communication account;
determining that the first country is a homo-ethnicity country based on a census data;
determining one or more popular account holder names associated with the first country;
assigning one or more dominant ethnicities associated with the first country to the one or more popular account holder names;
establishing a classifier model based on a machine learning algorithm, using, as training data, the one or more dominant ethnicities and one or more features of the first country derived from the word embedding; and
classifying a name into a classification using the classifier model, at least some of the method performed using a computer processor.
2. The computer-implemented method of claim 1 , comprising:
transmitting a communication to one or more electronic communication accounts based on the classification of the name.
3. The computer-implemented method of claim 1 , comprising:
disseminating targeted information, comprising at least one of an advertisement, a social media post, or news, based on the classification of the name.
4. The computer-implemented method of claim 1 , comprising:
generating the training data based on one or more contacts lists of one or more electronic communication accounts.
5. The computer-implemented method of claim 1 , comprising:
generating the training data based on one or more communication logs of one or more electronic communication accounts.
6. The computer-implemented method of claim 5 , comprising:
processing the one or more communication logs to extract one or more manners in which one or more users communicate.
7. The computer-implemented method of claim 1 , comprising:
generating the training data based on one or more contacts lists of one or more electronic communication accounts and one or more communication logs of the one or more electronic communication accounts.
8. A computing device comprising:
a computer processor; and
memory comprising instructions that when executed by the computer processor perform operations comprising:
extracting a first account holder name from a first electronic communication account;
extracting a second account holder name from a second electronic communication account;
extracting first contact names from a first contact list of the first electronic communication account;
extracting second contact names from a second contact list of the second electronic communication account;
sorting the first contact names in the first contact list to generate first sorted contact names;
sorting the second contact names in the second contact list to generate second sorted contact names;
creating, from the first sorted contact names and the second sorted contact names, a document comprising a plurality of words representing the first account holder name, the second account holder name, the first contact names and the second contact names;
generating a word embedding of a defined number of dimensions from the document for one or more words of the plurality of words;
identifying a first country corresponding to a first registration country for the first electronic communication account;
identifying a second country corresponding to a second registration country for the second electronic communication account;
determining that the first country is a homo-ethnicity country based on a census data;
determining one or more popular account holder names associated with the first country;
assigning one or more dominant ethnicities associated with the first country to the one or more popular account holder names;
establishing a classifier model based on a machine learning algorithm, using, as training data, the one or more dominant ethnicities and one or more features of the first country derived from the word embedding; and
classifying a name into a classification using the classifier model.
9. The computing device of claim 8 , the operations comprising:
transmitting a communication to one or more electronic communication accounts based on the classification of the name.
10. The computing device of claim 8 , the operations comprising:
disseminating targeted information, comprising at least one of an advertisement, a social media post, or news, based on the classification of the name.
11. The computing device of claim 8 , the operations comprising:
generating the training data based on one or more contacts lists of one or more electronic communication accounts.
12. The computing device of claim 8 , the operations comprising:
generating the training data based on one or more communication logs of one or more electronic communication accounts.
13. The computing device of claim 12 , the operations comprising:
processing the one or more communication logs to extract one or more manners in which one or more users communicate.
14. The computing device of claim 8 , the operations comprising:
generating the training data based on one or more contacts lists of one or more electronic communication accounts and one or more communication logs of the one or more electronic communication accounts.
15. Non-transitory storage comprising instructions that when executed perform operations comprising:
extracting a first account holder name from a first electronic communication account;
extracting a second account holder name from a second electronic communication account;
extracting first contact names from a first contact list of the first electronic communication account;
extracting second contact names from a second contact list of the second electronic communication account;
sorting the first contact names in the first contact list to generate first sorted contact names;
sorting the second contact names in the second contact list to generate second sorted contact names;
creating, from the first sorted contact names and the second sorted contact names, a document comprising a plurality of words representing the first account holder name, the second account holder name, the first contact names and the second contact names;
generating a word embedding of a defined number of dimensions from the document for one or more words of the plurality of words;
identifying a first country corresponding to a first registration country for the first electronic communication account;
identifying a second country corresponding to a second registration country for the second electronic communication account;
determining that the first country is a homo-ethnicity country based on a census data;
determining one or more popular account holder names associated with the first country;
assigning one or more dominant ethnicities associated with the first country to the one or more popular account holder names;
establishing a classifier model based on a machine learning algorithm, using, as training data, the one or more dominant ethnicities and one or more features of the first country derived from the word embedding; and
classifying a name into a classification using the classifier model.
16. The non-transitory storage of claim 15 , comprising:
transmitting a communication to one or more electronic communication accounts based on the classification of the name.
17. The non-transitory storage of claim 15 , comprising:
disseminating targeted information, comprising at least one of an advertisement, a social media post, or news, based on the classification of the name.
18. The non-transitory storage of claim 15 , comprising:
generating the training data based on one or more contacts lists of one or more electronic communication accounts.
19. The non-transitory storage of claim 15 , comprising:
generating the training data based on one or more communication logs of one or more electronic communication accounts.
20. The non-transitory storage of claim 19 , comprising:
processing the one or more communication logs to extract one or more manners in which one or more users communicate.