DVI — Freight Graffiti Research

🏠 Home (Intro)
📘 Methodology Paper
📊 Data Mining Tasks & Hashtags (/hashtags)

DVI — Freight Graffiti Research

🏠 Home (Intro)
📘 Methodology Paper
📊 Data Mining Tasks & Hashtags (/hashtags)

🧠 Multimodal AI Network Graph Analysis

Seed Node: freightgraffiti | Task MUID: freightgraffiti_0_hashtagRecent_30_be661c3d

Dashboard 📋 Task Details & Media Data

Mining Type: hashtagRecentPosts: 32Hashtags: 1Users: 0Inferences: 0

Created: 2024-04-08 06:00:00

Multi-Layer AI Network Graph Analysis

Interactive Graph Actions

Graphology.js ForceAtlas Layout — Parameters & Filters

Network will include:

Standard

Text_AI_Dict

Text_AI_OOV

Image_AI

Select the data used to feed the graph. Standard [hashtags, posts, users]. Text_AI: Hashtags classified by [Northamerica city list, Graffiti terms, Railroad terms]. Image_AI: Post classified by graffiti types [tag, wildstyle, 3D, monikers, bomba (throwup)]

Node Minimum Degree (0)Minimum entrance degree of node, smaller will be deleted from graph. Default: 0

Should we clean not related text_ai_entities to term "graffiti", numbers only and non latin chars?

true

false

Gravity (1)

Should the node's sizes be taken into account?

true

false

Use Barnes-Hut approximation O(n·log n)?

true

false

Barnes-Hut approximation theta (0.5)

outboundAttractionDistribution

true

false

Use Noack's LinLog model?

true

false

scalingRatio (1)

slowDown (1)

strongGravityMode

true

false

nodeFixedSize

true

false

initialLayout

random

circlepack

autoGravityScale

auto

manual

nodeReducerDepth *

selectedNeighbors (1-hop)

selectedNeighborsNeighbors (2-hop)

Iterations (133)

Building AI multi-layer graph network...

Click History

Click any node on graph...

🎓 Research ProjectUniversidad de GuadalajaraMaster in Communication Studies, 2024

Visualization of a Hypertextual Interaction Field in Network Form Using Computational Processes. Case Study: Freight Train Graffiti in North America 🔗

Developed by A. J. Ramírez Abundis

🔍 Researcher🌐 Ethnography🐍 idmb Mining Bot 🔗👁️ ML Vision (ResNet) 🔗🔤 spaCy Custom NER 🔗⚛️ DVI Web Interface 🔗

Master's Thesis in Communication Studies at Universidad de Guadalajara. This pioneering study explored the subcultural phenomenon of freight train graffiti and the structuring of a hidden transnational community of practice along North American railway tracks.

Methodologically, field and digital ethnography were combined with automated Instagram hashtag data mining through the development of the idmb bot 🔗. Hypertextual interaction metadata was modeled in complex network graphs using ForceAtlas layouts with SigmaJS and Graphology to analyze style propagation and recognition flows. Additionally, a CNN classifier (ResNet 🔗) was trained to recognize graffiti styles (monikers, tags, throw-ups). The conceptual framework integrated Jenkins' participatory culture and Thompson's social mediation theories.

📖 APA Citation

Ramírez Abundis, A. J. (2024). Visualization of a Hypertextual Interaction Field in Network Form Using Computational Processes. Case Study: Freight Train Graffiti in North America (Master's thesis). Universidad de Guadalajara, Mexico.

🌐 Read Full Thesis at abundis.com.mx

© 2026 DVI - Data Visualization Interface | Developed by Javier Abundis