This is an ongoing effort to both collect and analyze literature related to Industrial Symbiosis. To analyze the literature, we use Topic Modelling (video), which examines statistical properties of the text in order to determine the topics that are present.
Topics represent collections of words that frequently co-occur within a document, and by looking at the collection of words in a single topic, you can often deduce that the general theme is. For example, if you see in a topic the words "industrial, ecology, nature, metaphor, model, natural, ecosystem, analogy", then this topic is a collection of documents that are talking about Industrial Ecology and how it uses the metaphor or analogy of nature and ecosystems.
The visualization is interactive and allows you to explore several things:
- The collection of topics
- The collection of documents analyzed
- How closely topics are related to each other - based on a principal component analysis of the topics
- How the prevalence of topics has changed over time
- Which topics a word is prevalent in - this can show how one word might be used in different contexts
- Which words are relevant for the topics
Topics including specific words:
-
Evaluating EIPs/IS:
-
General Topics
-
Circular - Circular Economy
-
Organizations
-
Philosophy
- Metaphor - Nature as a metaphor for EIP/IS
-
Tools/Approaches
- Agent - Agent Based Models
- Fuzzy - Fuzzy Programming
- LCA - Life Cycle Assessment
- Matching - Match-making tools for linking industries
- Optimization / Optimisation
-
Energy Related:
-
Flows/Substances:
-
Locations:
-
Eco-Industrial Parks:
In the literature overview you can search by author, year, etc.
The source data used in the analysis below is from the file IndustrialSymbiosis.bib, which is stored using the BibTeX format. This is an open standard and can be read by most literature reference managers.
Entries in IndustrialSymbiosis.bib containing file = {Full Text PDF: indicate that we have the actual pdf of the article and that this is used in the analysis. Otherwise, only the title, abstract and keywords are used.
The visualization uses the great dfrtopics R package by Andrew Goldstone which in turn uses MALLET (MAchine Learning for LanguagE Toolkit) to perform the actual Topic Modelling.