Drafts by Abhishek Pandit
Strategic self-representation assumes centrestage in online dating,
which has grown exponentially... more Strategic self-representation assumes centrestage in online dating,
which has grown exponentially in popularity over the last decade.
With comparisons of dating portals to virtual marketplaces, users face
the challenge of honoring newfound norms while remaining memorable
in a sea of competitors. By juxtaposing theories of social psychology,
evolutionary biology and linguistic anthropology against a dataset of
18,000 users of a popular online dating site we explore two core ques-
tions. Firstly, has a social consensus emerged among heterosexual,
single American men around the constituents of an attractive dating
profile essay? Secondly, in what specific ways do these norms vary for
different demographics, especially where these users perceive a need to
either highlight or downplay certain attributes (such as height, weight,
race and education), based on their perceived role in attraction? We
explore topic modeling to identify differences in the proportion of var-
ious themes in each essay, before switching to confirmatory analysis
through structural topic models. We nd that extroversion and open-
ness emerge as preferred traits across the board, while black users'
pro les diverge from that of other groups along almost every topic.
Nonetheless, through both topic models and abstract vector space
representations, we find remarkable homogeneity in self-representation
across most demographics even in a setting as competitive as modern-
day online dating.
Uploads
Drafts by Abhishek Pandit
which has grown exponentially in popularity over the last decade.
With comparisons of dating portals to virtual marketplaces, users face
the challenge of honoring newfound norms while remaining memorable
in a sea of competitors. By juxtaposing theories of social psychology,
evolutionary biology and linguistic anthropology against a dataset of
18,000 users of a popular online dating site we explore two core ques-
tions. Firstly, has a social consensus emerged among heterosexual,
single American men around the constituents of an attractive dating
profile essay? Secondly, in what specific ways do these norms vary for
different demographics, especially where these users perceive a need to
either highlight or downplay certain attributes (such as height, weight,
race and education), based on their perceived role in attraction? We
explore topic modeling to identify differences in the proportion of var-
ious themes in each essay, before switching to confirmatory analysis
through structural topic models. We nd that extroversion and open-
ness emerge as preferred traits across the board, while black users'
pro les diverge from that of other groups along almost every topic.
Nonetheless, through both topic models and abstract vector space
representations, we find remarkable homogeneity in self-representation
across most demographics even in a setting as competitive as modern-
day online dating.
which has grown exponentially in popularity over the last decade.
With comparisons of dating portals to virtual marketplaces, users face
the challenge of honoring newfound norms while remaining memorable
in a sea of competitors. By juxtaposing theories of social psychology,
evolutionary biology and linguistic anthropology against a dataset of
18,000 users of a popular online dating site we explore two core ques-
tions. Firstly, has a social consensus emerged among heterosexual,
single American men around the constituents of an attractive dating
profile essay? Secondly, in what specific ways do these norms vary for
different demographics, especially where these users perceive a need to
either highlight or downplay certain attributes (such as height, weight,
race and education), based on their perceived role in attraction? We
explore topic modeling to identify differences in the proportion of var-
ious themes in each essay, before switching to confirmatory analysis
through structural topic models. We nd that extroversion and open-
ness emerge as preferred traits across the board, while black users'
pro les diverge from that of other groups along almost every topic.
Nonetheless, through both topic models and abstract vector space
representations, we find remarkable homogeneity in self-representation
across most demographics even in a setting as competitive as modern-
day online dating.