Extracting patterns from Twitter to promote biking

Das, Subasish; Dutta, Anandi K; Medina, Gabriella; Minjares-Kyle, Lisa; Elgart, Zachary · 2018 · IATSS Research

DOI: 10.1016/j.iatssr.2018.09.002

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of understanding the motivations and sentiments of bicycle commuters, a growing demographic in the United States. While traditional methods like surveys and crash data analysis provide limited insights due to resource constraints and lack of motivational context, social media mining offers a scalable alternative. The research aims to develop a framework for extracting knowledge from unstructured Twitter data to help transportation planners and decision-makers promote biking effectively. The researchers collected Twitter data associated with bike commuting hashtags over an eight-year period (2009–2016). They utilized a comprehensive list of hashtags, including #biketowork, #bikecommute, and variations involving "cycle" and "bicycle," to capture relevant conversations. Using open-source R packages, they gathered approximately 87,000 tweets, which were cleaned to remove spam, bot accounts, and redundant data, resulting in a final dataset of 80,563 tweets. The methodology involved natural language processing (NLP) techniques, including text mining, sentiment analysis, polarity scoring, and network analysis, to interpret the unstructured text and identify patterns in user interactions. The findings reveal distinct temporal and thematic patterns in bike commuting discourse. Tweet frequency peaked in May, correlating with National Bike Month, and was highest on weekdays, particularly Fridays. Text mining identified "day," "morning," and "today" as the most frequent terms, alongside positive descriptors like "beautiful" and "happy." However, weather-related terms such as "cold," "weather," and "rain" were also prominent, indicating that weather significantly influences biking behavior and sentiment. Sentiment analysis showed that the general attitude toward biking is positive, with polarity scores indicating increased positivity in recent years. Conversely, negative sentiments were primarily associated with adverse weather, crime, and other challenges. Network analysis of 26,102 unique interactions revealed that the majority of tweets were single interactions, with a small fraction involving multiple users, highlighting the structure of information sharing within the community. The significance of this study lies in its provision of a replicable framework for using social media analytics in transportation research. By demonstrating that Twitter data can effectively capture the sentiments, motivations, and challenges of bike commuters, the research offers a low-cost, high-volume alternative to traditional surveys. The insights gained, particularly regarding the impact of weather and seasonal trends, can assist planners in designing targeted campaigns and infrastructure improvements to encourage non-motorized travel. This approach fills a gap in transportation literature by leveraging microblogging conversations to understand public opinion and behavior at a scale previously difficult to achieve.

Key finding

Analysis of Twitter data reveals that bike commuting sentiment is generally positive but significantly influenced by weather conditions and seasonal patterns, with peak activity occurring in May.

Methodology

dataset

Sample size: 80563

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• cyclist safety