Matching Self Presentaion in Internet Dating Sites to Consumer Preferences: An Innovative Matching Algorithm

This study presents an innovative "Matching Algorithm" to match self presentation to consumer preferences in internet dating sites using data mining and machine learning techniques. The study is designed from 2 parts: The first part examines the correlation between the presentation characteristics of man and women in social networks vs. the response rate using several hypotheses. Results show that there is a strong correlation between the way man and woman presents themselves in social networks (such as "FaceBook") especially in the range of ages 18-55 (average age is 25.91). In addition, there is a strong positive correlation between the desire of man and woman to develop a romantic relationship between them trough social networks. As such, the more the user desires to achieve a "Real" relationship that may lead to a serious long term relationship, the more he/she uses the social network as an application to achieve their objectives. In the second part the author used data mining and machine learning techniques (Decision trees and Genetic Algorithms) to predict which personal attributes may influence the response rate of the other side's (In this paper only Decision trees – J48 algorithm results will be shown). Results show that some attributes (characteristics) related to personal presentation and education background are critical to achieve a positive response from the other side.