As mentioned in our introductory slides (Introduction [1/2]: General introduction), Persagen is committed to universal social justice and equality; ecosocialism; human rights. We track and curate wealth, power, and influence affecting those issues.

Persagen.org seeks to deliver high-quality, factual information for the benefit of humanity.

How do we do that? Data curation and analysis at Persagen is a multi-faceted approach.

Sources

Textual knowledge store

Ontology

Knowledge graph

As mentioned in our introductory slides (Introduction [1/2]: General introduction), Persagen is committed to universal social justice and equality; ecosocialism; human rights. We track and curate wealth, power, and influence affecting those issues.

Persagen.org seeks to deliver high-quality, factual information for the benefit of humanity.

How do we do that? Data curation and analysis at Persagen is a multi-faceted approach.

Sources

• The meticulous and rigorous attention to the selection of content ingested by Persagen is described in our Sources page.

• The overarching goal is the curation of factual, reliable information.

Textual knowledge store

• Textual data is curated in HTML files, which are indexed (according to a schema, capturing metadata) in Apache Solr.

• Pre-processing steps sanitizes the data (e.g. removing trackers), while adding important, missing information (web links; ...).

• The power of Pg-Solr (our interface to Solr) is evident in the responsiveness and flexibility of searches, e.g. Leonard Leo

Foundation for Government Accountability

Ontology

• All documents are categorized in a well-defined, grounded hierarchy - showing context and relationships.

• As well as grouping relevant news items and other information to ontological entries (HTML web pages), the ontology also serves as a handy glossary and quick-reference guide.

Knowledge graph

• Like a table of contents or a dictionary, an ontology is useful for broadly organizing and displaying knowledge. However, cross-references and other nuanced (implicit) relationships among data are better served in a knowledge graph.

• Persagen's knowledge graph explicitly maps the relationships among all curated data, providing easy-to-understand visual representations of the data, facilitating enhanced searching and understanding of the relationships within the data.

• The information and knowledge contained within Persagen (Pg-Solr search; Eureka! ontology; Leibniz knowledge graph - in progress) is comprehensive and authoritative.

• Earlier in the project (solo effort) I experimented with the addition of content to HTML files and the styling (content, links, etc.) on those pages. That remains the preferred approach.

• However, to keep up with the onslaught of news related to Persagen's mission (tracking wealth, power, influence) I have defaulted to adding that content to my FAQ § What is an ontology? (the raw text file is available; CAUTION - large file: 62,675 lines; 35MB) and converting those entries to temporary HTML pages via a bespoke script. Those HTML pages are then indexed in Pg-Solr, making all content on Persagen available to search. (Because of those time / financial limitations, the formatting may be less than desirable, but the information is there.)

Financial support to Persagen allows

• better monitoring of sources for new and relevant information;

• curation and addition of high-quality content;

• improved interface;

• improved visualizations; and,

• construction of the knowledge graph (Leibniz), in support of the above.

What is an ontology? 🤔

A brief overview of ontologies is provided on our frequently asked questions page, FAQ § What is an ontology?, and explained in greater detail in our blog entry.

To reiterate:

• An ontology is a system for organizing (classifying; categorizing) information. For example, entries in a dictionary or an encyclopedia are sorted alphabetically - the ontological structure is the alphabet (A, B, C, ..., X, Y, Z).

• However - although ordered, facilitating easy browsing - that ontology is both shallow (depth is the alphabetical sorting of the first word or phrase) and arbitrary (the relation of entries to adjacent entries is largely random).

• In contrast, Persagen.org uses a grounded, hierarchical ontology. The root of the ontology encapsulates the universe and everything in it (real or imagined) - e.g. "Root - Nature - Earth - Countries - Canada".

• Importantly, everything in the ontology may be uniquely identified, and the ontological placement provides information about relationships among entries.

• A grounded ontology forms a type of graph structure - hierarchical "tree" - with paths branching out from the "Root" node, terminating in "leaf nodes" (the actual entry) - which may be a short description, a web page, etc.

• Ontological entries - which serve as repositories for relevant information - are easily cross-referenced.

• While fundamentally important for categorizing knowledge, a more nuanced mapping of relationships among the information contained within an ontology requires a higher-dimensional structure called a knowledge graph - which permits both the querying and visualization of those relationships - facilitating knowledge discovery.

• Lastly, in addition to providing information and collating relevant information and cross-references, summaries in the leaves (terminal nodes) of the ontology double as a glossary and quick-reference guide.

An integral component of the Persagen ecosystem is the Persagen knowledge graph (Leibniz , named in honor of Gottfried Wilhelm Leibniz).

In knowledge representation and reasoning, a knowledge graph is a knowledge base that uses a graph-structured data model (relationships) to integrate data.

Knowledge graphs are often used to store interlinked descriptions of entities - objects, events, situations or abstract concepts - while also encoding the semantics underlying the used terminology.

Knowledge graphs are structured knowledge models that explicitly store rich factual knowledge.

Since the development of the Semantic Web, knowledge graphs are often associated with linked open data projects, focusing on the connections between concepts and entities. They are also prominently associated with and used by search engines such as Google, Bing, ...; knowledge-engines (e.g. Persagen! 😀) and question-answering services such as WolframAlpha, Apple's Siri, and Amazon Alexa; and social networks such as LinkedIn and Facebook.

Like the ontology (Eureka!) this work is under active development (limitations of time, resources but more substantially the data structure and implementation).

Ontological data is perhaps best stored as JSON - used in the D3.js visualization a couple of slides back. However, the Persagen ontology Eureka is very large (~31k entries, 2023-06) and deep (many levels, for some entries) - complicating the loading and visualization of those data as shown in the D3.js visualization. There are solutions: e.g. on-demand loading of those date (e.g. Apache Arrow  |  reading JSON files, perhaps).

Likewise, the data format, storage and visualization of knowledge graph data is under investigation (these are all data science / engineering issues). Storage perhaps as JSONB in PostgreSQL (I have experience with MySQL and PostgreSQL). The data visualization might employ NetworkX, Cytoscape. or some other approach leading to a searchable semantic property graph). I've all of the preceding platforms and more; I also worked with Neo4j (a graph database); the commercial license is cost-prohibitive, and I found it to be sluggish on larger datasets.

• Another possibility is Apache AGE, although development on that project appears slow, and development of the Apache-Age Viewer appears to have stalled, unfortunately. Apache AGE is attractive as it natively integrates PostgreSQL ...

Note: although I (Persagen creator/owner Victoria Stuart; then as G.R. Stuart) have prior published experience building biomedical knowledge graphs in Cytoscape (and, a JavaScript version of Cytoscape is available), I likely will use an alternative (bespoke) approach.

• Stuart G.R. et al. (2009) Construction and application of a protein and genetic interaction network (yeast interactome). Nucleic Acids Research. 37(7): e54. DOI: 10.1093/nar/gkp140.  |  pdf

• Stuart G.R. et al. (2009) Transcriptional response to mitochondrial NADH kinase deficiency in Saccharomyces cerevisiae. Mitochondrion. 9(3): 211-221. DOI: 10.1016/j.mito.2009.02.004.  |  pdf

We hope this brief introduction has been helpful.

In summary, Persagen stands above all existing knowledge sources for its:

• Accessibility to key personnel (points of contact).

• Attention to detail.

• Careful selection and curation of content.

• Processing and storing of that information.

• Tools to rapidly search those data.

• Tools to visualize and understand those data.

No other site provides those resources - notably also assembled within a socially responsible philosophical framework.

---

[toggle] A note from the owner/creator Dr. Victoria A. Stuart, Ph.D.

• I place my name and reputation on this website and its contents (read about me here). It is a work of passion. It matters.

• I am trans, 62 (in 2023), over-educated and unemployed. I don't need or want much money - enough to survive, sustainably (food; rent; ...), with excess monies reinvested in Persagen and its mission.

• If it matters (you may want to put your money elsewhere) I am white, but abhorrent of all forms of discrimination, prejudice, and racism - including its structural and systemic manifestations.

Accordingly, I advocate for Asian, Black, Brown, Hispanic, Indigenous, and other marginalized groups (LGBT+; ...), where-ever you are. 🙂

• I am the main point of contact; if you have any site-related issues or concerns, please do not hesitate to contact me. ✉

mail@persagen.org (Attention: Victoria).

• For general questions, please contact Carm at info@persagen.org (Attention: Carm).

---

[End of slides]