Information Technology Reference
In-Depth Information
3.3 Relational User Attribute Inference Problem Formulation
User Attributes
We use user attribute to describe a type of user property in social
media networks. In this work, we consider six types of user attributes includ-
ing three biographic attributes—
age, gender
,
relationship
, and three personaliza-
tion attributes—
occupation, interest
,
sentiment orientation
. The attribute values are
defined manually based on a comprehensive study of Google
data and a survey of
previous work on user attribute inference [
13
,
28
,
38
]. Table
3.2
presents the mean-
ing of user attribute values.
5
The six types of attributes definition are described as
follows:
•
+
Gender
. Gender is a binary valued attribute. We use gender to describe whether a
user is male or female.
•
Age
. Age is a real valued attribute. Given the general lack of ground truth for user
age, exact age inference is impossible in social networks. We conduct a detailed
investigation and observation on Google
+
users. Generally, Google
+
users can
Table 3.2
User attribute definition
Attribute name
Attribute values
Gender
1 Male; 2 Female
Age
1 Young(
≤
30); 2 Elder(
≥
30)
Relationship 1 Unmarried; 2 Married
Occupation 1 Student(St); 2 Information Technology Person (IT), Software Engineer,
Geek; 3 Entertainer, Musician, Actor, Comedian, Model, TV show host;
4 Writer, Journalist, Blogger, Editor, TV news host, Critics Lawyer;
5 Politician; 6 Sports star, Athlete; 7 Business man, Economist,
Entrepreneur, Market strategist, Financiers; 8 Scientist, Professional,
Researcher, Expert; 9 Photographer Traveler; 10 Doctor, Dentist,
Pharmacist, Beautician; 11 Chef, Eater, Cook; 12 Engineer, Specialist,
Designer; 13 Teacher; 14 Artist, Religious people, Culture Writer,
Designer, Author, Critic; 15 Other
Interest 1 Technology, Information, Internet; 2 News, Politics,military, Society;
3 Economy, Business Manage Strategy; 4 Entertainment, Music, Movie,
Fashion; 5 Photography, Travel; 6 Food&Drink; 7 Daily things,
Lives life living, Fun interest, Personal Stuff; 8 Sports, Exercise,
Body-Building; 9 Thinker, ideas religion culture literature art; 10 Health,
Medical care, Treatment, Makeup; 11 Science, Knowledge; 12 Other
Sentiment orientation 1 Positive (fantastic, great, elated, bouncy, jubilant, excited, cheerful,
ecstatic); 2 Negative (annoyed, aggravated, bad, pain, embarrassed,
bored, anxious, crazy, depressed, scared, sick, angry, sad, score);
3 Neutral (normal, awake, calm, working, blank, report, news, fact)
5
In the following sections of this chapter, we will mix the usage of “attribute” and “attribute value”
when no ambiguity is caused.
