Data mining is widely used in areas such as healthcare, finance, retail, and education. Recent studies indicate that it is also proving a useful tool to study social behaviors. Let’s look at some interesting examples of this.
Data Mining to Analyze Behavior on Online Dating Sites
Online dating is a huge business these days and generates billions of dollars in revenue. Out of the 54 million single residents in the U.S, 40 million use online dating websites such as match.com and eHarmony. Forty million is not a small figure and there is a lot of curiosity in understanding the behavior of the members of these sites so that a system can be designed to create a successful match.
A report published in the MIT Technology Review in January described the efforts of a group of researchers at the University of Massachusetts Lowell to study the behavior of 200,000 members of a Chinese online dating site using data mining techniques. Each person’s profile includes details such as age, gender, location, body type, occupation, marital status, children status, and so on. It also has information on the dates of all the messages they sent during an eight week period in 2011, and also the receiver of the message, and whether they responded. Their study came up with interesting findings:
- Males tend to look for younger females
- Females are concerned about the socioeconomic status of males, including income and education level
- Men send far more messages than women but get fewer replies
- A large number of messages are sent to or replied to users whose qualities do not match the sender or receiver’s stated preferences, with women more likely to deviate
- People’s choices are essentially random when it comes to certain categories such as height, education, and so on.
- When choosing partners, people’s actual behavior differs significantly from their stated tastes and preferences.
As the report points out, such information is crucial when it comes to designing the questionnaires posted on dating websites as well as the algorithms that connects one person to another. In other words, data mining strategies used the right way can help generate information that can be used to improve the services offered by these websites.
Mining Data to Understand Cultural Differences
Another report published in the MIT Technology Review in April reveals how mining data can be helpful in studying cultural differences across the globe. However, studying values and their impact on social and political life in different societies in various countries has always been a challenging task. Researchers at the Universidade Federal de Minas Gerais in Brazil have simplified this by analyzing cultural differences in different parts of the world on the basis of data generated by check-ins on Foursquare, the location-based social network. This made data capture easy, quick and affordable.
Cultural analysis of a particular location often starts with the evaluation of eating and drinking habits as they act as important indicators in determining the habits and behaviors of a society. People’s food and drink preferences change with geographical location and time of day. Foursquare is an ideal source to mine such information as users “check in” by indicating when they have reached a particular location that could be related to eating and drinking as well as to other activities such as entertainment, sport and more.
These researchers study only food and drink preferences of individuals and the manner in which these tastes change according to time of day and geographical location. Comparisons of this information will help them understand individual preferences from different countries and how they are similar or different from each other.
The team mined information related to food or drink from tweets containing Foursquare check-ins, URLs to the Foursquare website with information about each venue. They got check-ins related to drink, check-ins related to fast food, and check-ins relating to ordinary restaurant food. These classes are then divided into subcategories. Mining this information allowed them to compare:
- eating and drinking times in different countries both during the week and at the weekend
- choices of restaurants, fast food habits and drinking habits by continent and country
- eating and drinking habits in different cities across the world
The results of the study are providing fascinating insight into humanity’s differing habits.
The applications of data mining are on the rise. However, analyzing big data can be a challenging task for businesses as they need to focus on their core tasks. The best solution in such situations is to partner with a professional data mining company that has the resources to accomplish the task efficiently.
