i2Insights as a repository

By Gabriele Bammer

Gabriele Bammer (biography)

How is Integration and Implementation Insights (i2Insights) shaping up as a repository of resources useful for tackling complex societal and environmental problems?

i2Insights has two major purposes:

  1. connecting a community of researchers to each other, and
  2. building a repository or knowledge bank of resources.

i2Insights has set out to achieve both purposes using the format of blog, with short, easy-to-read contributions from researchers located anywhere in the world, and with encouragement to peers to comment. We have sought to summarise these purposes in the tagline for i2S:

A community blog providing research resources for understanding and acting on complex real-world problems

Types of resources

To date, i2Insights has published 425 blog posts, which can be categorized as follows (number of blog posts in brackets, with any particular blog post able to fall under more than one category):

Topics covered

When it comes to topics covered, the overarching consideration is that research dealing with any complex societal or environmental problem involves:

  1. developing a more comprehensive understanding of the problem, both what is known and what is not known
  2. generating ideas about addressing the problem, including what may and may not work
  3. supporting improved policy and practice responses to the problem by government, business and civil society.

The 14 specific topics that i2Insights therefore deals with can be grouped into 4 clusters (number of blog posts in brackets, with any particular blog post able to fall under more than one category):

Diversity, teamwork, stakeholder engagement and integration

Decision-making, research implementation, change and communication

Systems, context and unknowns

Evaluation, education and institutionalisation

  • evaluation to improve effectiveness in each topic (27)
  • education to develop expertise in each topic (31)
  • institutionalisation to ensure that the expertise is recognised, widely adopted and rewarded (52).

i2Insights aims to publish both long-established and newly-developed concepts, methods and other resources on these topics.

Communities of practice

i2Insights aims to be a repository for resources from researchers in established communities of practice (such as transdisciplinarity, systems thinking, action research), as well as from researchers who do not necessarily identify with such communities.

The number of contributions linked to specific communities of practice are:

i2Insights is a project supporting the formation of Integration and Implementation Sciences (i2S), a discipline which aims to provide a way of sharing insights and resources among all those tackling complex societal and environmental problems. This includes those in 22 different communities of practice, some of which are mentioned above – for details see https://i2s.anu.edu.au/what-i2s/faq/where-i2s-fits. Not all communities of practice are yet represented in i2Insights.

There are also 18 blog posts about i2S and you can read more about the discipline on our companion website at: https://i2s.anu.edu.au/.

Achieving global reach

i2Insights aims to attract a global authorship and readership. To end March 2022, there were 530 authors from 46 countries. The bulk of contributions still come from a small number of countries (USA, Australia, UK, Germany, Switzerland and Canada). We are keen to encourage contributions from other countries.

Since the inception of i2Insights, contributions have been viewed in 188 of the 193 countries that are members of the United Nations. There have been more than 100,000 views from USA, Australia and UK, with more than 10,000 views from Canada, India, South Africa, Philippines, Germany, Switzerland, Netherlands, New Zealand and Pakistan.

More statistics can be found at https://i2insights.org/about/i2insights-statistics/.

Index and advanced search 

Key to any repository is an effective index and advanced search. All blog posts are indexed using a consistent set of terms, which includes the categories above for resources and topics covered, along with hundreds of more specific tags, ranging from academic career development, through hedging, negotiation and tacit knowledge, to yin-yang. There are also tags for languages other than English in which blog posts are also published. To date, contributions have been published in Arabic (1 blog post), Chinese (1), French (3), Japanese (1), Portuguese (1) and Spanish (10).  We also provide easy access to Google Translate in the menu bar.

Help i2Insights improve

Although our resources are limited, we are always interested in suggestions for improvement and general feedback. In particular, what would make i2Insights a more useful repository of resources for you? If you are not already a contributor, what would encourage you to become one?

We also welcome new i2Insights Ambassadors, who help the blog expand its reach so that the exchange, discussion, network building and repository become truly global and comprehensive efforts. i2Insights Ambassadors help us achieve our aims of having a vibrant community of contributors in every country and having extensive treatment of all relevant topics.

Biography: Gabriele Bammer PhD is Professor of Integration and Implementation Sciences (i2S) at the National Centre for Epidemiology and Population Health at The Australian National University in Canberra. i2S provides theory and methods for tackling complex societal and environmental problems, especially for synthesis of disciplinary and stakeholder knowledge, understanding and managing diverse unknowns, and providing integrated research support for policy and practice change.

21 thoughts on “i2Insights as a repository”

  1. Thanks for the synthesis of the blogs. Reading them one by one as they are published does not give me an appreciation of the breadth of i2S, which your Review provides. The content and context of i2S seems to be organic, continually emerging as authors see a link that others might miss. Mapping the process would be interesting as long as the danger maps bring of proscribing what’s in and what’s out can be avoided. I do use the blogs as a resource when planning interventions of specific complex issues as they have a capacity to catalyze my thinking to develop novel ways of approaching a new challenge.

      • Bruce raises interesting points, though I think there is a conversation still lurking about “danger of maps proscribing what’s in and what’s out”. That’s not an issue I’ve given much thought to and would like to hear more.

        Let me offer this hint: when I say “topic map” I am talking about a map of one or more territories, where the prime territory is defined by what’s in i2S but that’s not the whole story; i2S covers a massive range of topics, each an individual “location” in a topic map, with relationships connecting them, biographies describing those relationships, and so forth.

        In theory, a topic map is not the static object of a road map you used to take with you on long drives; it’s more like your GPS client, always dynamic,sensitive to changes in the territory.

        What is clear to me is this: a topic map is first generated – the way I prefer to do this – through harvesting tools which, in effect,harvest what’s in the database, using all manner of “AI” (artificial intelligence) magic-mostly simple and common “NLP” (natural language processing) tools – to learn what’s in the territory (the database). But that’s not the whole story. I see “humans in the loop” assisting in curation of the database. So, it’s a bit like punctuated evolution: there’s a huge burst (Cambrian explosion?) at first when the harvesting is going on, then bumps along the way as new information arrives in the database and curators perform whatever magic is needed. Evolutionary improvements to AI means more depth, breadth, and precision over time.

        How that could possibly proscribe “what’s out” remains, for me, a mystery. Eager minds want to learn…

        • Jack the mapping you are describing is beyond my knowledge. I was referring to the fact that a map’s boundaries tend to very quickly become fixed and the dynamic you mention occurs within that adopted boundary (often we are tempted to adjust the dynamic to enable the insight to fit inside the stated boundary). Whereas I’m experiencing i2S as having porous boundaries that are reshaping as unknown unknowns are exposed and explored. I suspect the issue we are reflecting on is one created by our use of language, which is a continuing challenge for any attempt to define i2S.

          • Bruce, thanks for that clarification (I think!).

            If one’s mental model is that of a road map, it clearly proscribes what’s inside, say, some town. No room for porous boundaries there. Some house is either in, or not in the city (except for those pesky ones which were built across lines – or lines moved).

            But, a topic map is not like that. It has no borders.

            What it has is what are known as “applications” – Topic Map Applications (TMAs), each of which specifically defines what it is mapping. For example, a TMA for People is going to be different from a TMA for Locations. But, they live together in the same topic map. As we grow a map, we can grow it by adding more topics to each TMA, or we can add more TMAs and grow those.

            In terms of your porous boundary notion, I’d have to say that a Topic Map doesn’t have boundaries at all. Sure, the curriculum associated with i2S prescribes what’s in, but over time, that is going to change; evolutionary epistemology happens.

            That’s the best I can come up with for the moment; here on the left coast of the colonies, we are about to start blowing up fireworks and I’m mostly exhausted.

            Final thought: the term topic map may be new to you, or not even in your wheel house, but the term “knowledge graph” seems to be a newly minted “common” term. A topic map is very much like what is called a knowledge graph, but a bit more complex. I did a piece in this venue about maps and mapping a few years back.

              • The two aspects of any knowledge cartography project which I think are crucial are these:
                * Knowledge Representation – the internals of the platform
                * UX (user experience) – making all that information useful – a truly tall issue.

                People frequently draw graphs as bubbles and lines. Those are really fun and appealing, right up until there are so many of them that you can’t find anything.

                People sometimes organize resources as indented outlines. But then, when some nodes are connected to others in ways that an indented outline doesn’t show well, you’ve got more UX issues.

                A simple way to present a topic is as a dashboard: primary content,e.g. descriptions, images, etc, as central, then typed links around the periphery. For instance, if the topic resides in a taxonomy, super class links at the top, subclass links (if any) at the bottom. Outbound relations on the right, inbound links on the left – that sort of thing.

                Interpretation should be left to the beholder; UX should make that possible.

                None of what I just said is trivial; I believe we still have a lot of exploration to do. No one single UX can satisfy all needs. As an example, https://debategraph.org/ is a structured conversation platform which can be viewed as bubbles and arcs, pages (akin to dashboards), or outline views.

              • I have just read Jack’s blog that your referenced Gabriele and I get the sense of where his comments are grounded. It does further raise the question of the language used in different arenas to describe similar concepts and some of the ‘crunch’ points when we integrate various disciplines. To a systems thinking practitioner a map (an image of a system) by definition can’t be without a boundary, we may not be sure where it should be drawn but it’s there somewhere. Dialogue Maps are often called Conversation Maps and are created to provide a database for emerging new insights about the issue in focus. Maybe there is a future blog in the exploration of ‘mapping’ as a tool in i2S discourses. Thanks Jack for your patient explanation I will pursue my study of the ideas you promote.

                • Thanks Bruce. I think the issue of boundaries is also an interesting discussion point, along with mapping as a tool and language differences.

                  I’d frame the boundary issue a little differently – because we can never consider the whole system, we have to set a boundary. In the case of a topic map, either the material to go into the map would run out or interpreting the map would become impossible, so at some point it would presumably reach its limits.

                  Thanks to you both for an enlightening discussion.

      • It’s a non-trivial project if you take topic mapping to its limits, something to discuss. It’s a software/IT project to go with the blog’s database.

        • Indeed. It’s taken us 2.5 years to do a first pass at a consistent way of indexing the blog posts and in light of what we have learnt, it’s ready for another big project of refinement in consultation with information science specialists. But it’s a great way of thinking more deeply about what Integration and Implementation Sciences is and can be. I expect the process of producing a topic map will be just as valuable as the outcome. And, of course, we’d need expert guidance / collaboration. This is not something we have the resources for now, but it is something to look for opportunities to do.

          • Always happy to help 🙂
            It’s a good sign that you take it as a serious information sciences project, as reflected in this blog resource. Everyone will learn from the exercise.

          • You like to use the term “integration” and I like to use the term “federation”. It’s a different conversation but I like to think of federation as a way to bring disparate information resources and concepts under one tent – BigTent – and organize them in such a way that relationships among them can be discovered. Frequently, two different concepts might be the same. Federation should be able to recognize that.

            A trivial example – the one I used in my PhD work – was this:
            * Statement A: CO2 causes climate change
            * Statement B: Climate change is caused by carbon dioxide.

            To English speakers, it’s “QED”: they say the same thing.
            A federation engine, by simply comparing strings, would not see that. We have to apply some NLP (natural language processing) to disparate resources to establish “meanings” (badly overused term) in order to perform federation processes. Topic Maps are about that.

            • I think I see what you are getting at, but I can’t see how to bridge the gulf in my understanding without a lot more time and effort than I can currently devote.

              It’s good to have had the conversation – always good to sow some seeds ready for the day when conditions are right for germination.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.