Computational Anthropology: A New and Upcoming Field

What is Computational Anthropology?

As technology begins to influence many spheres of our society, it has gradually begun to impact a variety of academic disciplines, including anthropology. More specifically, the emerging field of computational anthropology allows anthropologists to identify trends and insights about human behavior and migration that were previously undiscovered. By uncovering and analyzing vast expanses of public data, anthropologists and data scientists are able to make deeper and more accurate forecasts about human mobility patterns. The data gathered from these algorithms can potentially allow researchers to effectively model a disease's growth or simulate certain historical events.

Research Conducted by Microsoft

Research conducted by Microsoft in Beijing sought to answer the question: where is a user most likely to visit given their current location and recent pattern of migration. Answering such a question would imply analyzing the user's data (e.g. hometowns and locations recently visited) and coming up with a list of other locations that they are likely to visit based on the type of people who visited these locations in the past.

The researchers started with a hypothesis that claimed people who were living in a city (locals) had very different mobility patterns than tourists who were merely visiting. By stratifying users into local and nonlocal categories, researchers could improve their ability to predict where people would most likely visit.

Source: MIT Technology Review

To collect data, researchers begin with data from Jiepang.com. This site lets users record and document locations they have recently visited, allowing them to connect with friends in similar locations and find people who share similar interests.

The data points collected by the researchers were known as 'check-ins,' and the team collected over 1.3 million of them in five major Chinese cities. Once the data was collected, it was inputted into a predictive algorithm: 90 percent of the data was used to train the algorithm, and the remaining 10 percent was used to test it. The researchers then utilized this training dataset to grasp the mobility trends of locals and non-locals and the popularity levels of their destinations. Subsequently, they implemented this knowledge onto the test dataset to evaluate the efficacy of their algorithm in predicting the probable destinations of locals and non-locals.

As the researchers began to train their model further, they noticed that optimal results were produced when they analyzed the pattern of behavior of a particular individual and estimated the extent to which this person behaves like a local. This created a weighting scale called an indigenization coefficient that could be used to determine the mobility of patterns that someone was likely to follow in the future.

This could have a tremendous impact on the way that anthropologists study migration and the way immigrants and refugees assimilate into the local community. Sciences like computational anthropology hold immense potential to reshape our future and hopefully create more informed and inclusive societies.

Inspiration + Read More