In the spring of 2019, the L train in New York City will undergo a sustained
period of repairs. We can learn a lot about the value of transportation
infrastructure by observing the effects of this disruption.

The L train is a line in New York City’s subway system, controlled by the
Metropolitan Transportation Authority (MTA) and connects the far reaches of
Brooklyn and parts of Queens with Manhattan. It is among the busiest subway
lines in the city: according to the MTA, 225,000 people a day use the L Train to
travel between Brooklyn and Manhattan and 50,000 people a day use the service
solely in Manhattan.

The service delays and suspensions associated with the L train repair project
will usher in a period of sustained uncertainty for individuals and businesses
that rely on the transit line and alter travel patterns across the city. Who
adjusts their behavior to compensate and how they do so can tell us a lot about
the value of this piece of infrastructure across multiple scales to individuals,
businesses, communities, and the city and region..

We analyze geolocated Twitter data from from New York using natural language
processing methods such as sentiment analysis and topic modeling to learn about
the public opinion towards different aspects of the shutdown. We further perform
hot spot analysis to determine regions in which different topics or sentiments
tend to cluster.

First results give us an indication about which topics are being discussed more
emotionally than others and where significant clusters of these topics occur.
This information can help urban planners by showing what aspects of the shutdown
are important to the public and develop mitigation and communication strategies
accordingly.