Outis wrote: Mon Jan 20, 2025 8:36 pm
Qualitive and quantitive both provide value.
Also, don't be afraid to go off the beaten path and try things that you wouldn't expect to have any impact.
It's more difficult for map advocacy since the metrics for measuring impact are less clear. If selling a product or service then it's easy, you measure sales figures or enquiries, but what kind of metrics can be measured to detect if something is working?
Here's an initial approach to consider.
1. Define a goal and some metrics.
We should clearly define what success looks like such as a better understanding of maps in our target audience, community growth, a more tolerent approach to maps in our target audience etc.
Metrics could be things like this.
- Mailing list signups
- More community members in map forums
- More likes and shares of posts
- Website traffic up
- Survey results showing increased awareness
- Behavioural changes such as improved sentiment towards maps
2. Define presentation variables to modify
Medium - Articles, videos, podcasts, infographics, presentations, social media posts etc
Tone and Style - Formal vs Casual, emotional vs logical, storytelling vs academic, real life stories vs research etc
Target audience - Different demographics and psychographic groups. Academics, parents, People in their 20s, 50s, males, females, students, professionals, young adults etc
Channels: LinkedIn, Twitter, newsletters, mail drops, conferences, emails etc
Content angle: Focus on benefits, addressing common objections, providing evidence, real world examples etc.
3. Design experiments
Create a schedule to test differences, such as changing one variable per week.
Use control groups, so change a variable such as style but do so as an article and video, or same audience with a different tone.
4. Measure impact
- Website, social media analytics etc
- surveys and polls at intervals, say every 6 months
- focus group interviews
- Conversion tracking
5. Analyse results to work out what works and what doesn't
6. Iterate and scale out
- Refine approach based on learning
- Increase what works and eliminate what doesn't
- Try smaller variations to fine tune knowledge
This is the general approach taken to sell ideas as well as products and services. It sounds intense but it can start small and build out over time.
At first most things will fail but over time we gain a clear understanding of what works and what doesn't, what wins people over to our side, while winning supporters. Start small and be prepared for disappointments but that's normal, it weeds out the things that don't work and refines a powerful and effective strategy that does work, and that can spread and help the whole community.
4.
the article below is about BLM, but I guess could be MAP Liberation also....
"The article is dedicated to memetic practices of symbolic protest images in
online activism, focusing on net memes and net icons of protest. Considering
the particular affordances in the creation, distribution and reception of
memetic imagery in online activism, key research questions are: (1) what are
distinctive memetic practices in online activism of iconizing images on social
media platforms? (2) what are specific differences between net memes and
iconic imagery of protest in social media environments? and (3) how do
emblematic image events on the street and iconic imagery on the internet
interact in visual practices of online activists? After a theoretical discussion,
these questions will be addressed by means of a qualitative analysis of a
small sample of social media postings around civil rights activism (Black
Lives Matter and the Belarussian Democracy Movement). Starting from a
classic news icon originating from the protests against the Vietnam War
(‘Ultimate Confrontation’ by press photographer, Marc Riboud, in 1967), this
article analyses its recent memetic appropriations on digital platforms and
in performative protest actions on the street." -Net icons and memetic imagery of protest
in online activism
also..
"The term metaheuristic was coined by Fred Glover in his seminal
article ‘‘Future Paths for Integer Programming and Links to Artificial
Intelligence’’ (Glover, 1986). A metaheuristic can be seen as a methodology that includes master strategies capable of guiding the search for
the globally optimal solution. They are considered more complex and
efficient than simple heuristic algorithms because they explore areas
in the solution space that go beyond those explored by the simple
heuristics, which tend to focus on finding a single locally optimal
solution"
"In this paper, we will adopt the definition of Sörensen and Glover
(2013): ‘‘A metaheuristic is a high-level problem-independent algorithmic
framework that provides a set of guidelines or strategies to develop heuristic
optimization algorithms’’. In this definition, a metaheuristic is itself not
an algorithm (i.e., a precisely defined series of steps), but plays a looser
role as a more or less consistent set of high-level ideas that can be
used to develop a problem specific heuristic optimization algorithm"
"2. Constructive metaheuristics
As the name suggests, heuristics based on a constructive metaheuristic construct solutions from their constituting elements. These elements
depend on the model that is being solved. Examples include: the items
in a knapsack problem, the arcs between nodes in a routing problem,
the order of the tasks in a scheduling problem, etc. Generally, the
constructive process starts from an empty solution, i.e., a solution in
which the status of each solution element, either part of the solution or
not, is undefined" -Fifty years of metaheuristics
" Some
popular population-based metaheuristics include memetic algorithms
(MA), biased random key genetic algorithms (BRKGA), scatter search
(SS), and path relinking (PR). Each of these algorithms has its unique
characteristics, but they all share the fundamental principles of maintaining a population of candidate solutions and iteratively improving
them through inspired search and interaction mechanisms."
"4.1. Memetic algorithms
The general idea behind memetic algorithms is to exploit all possible
knowledge of the problem being solved inside the solution process. This
is also where the name ‘‘memetic’’ takes its roots. The knowledge can
take different forms, but always in the goal of favoring the balance between exploration and exploitation. These mechanisms are designed to
overcome the difficulties encountered by traditional genetic algorithms.
Moscato (1989) designed the memetic algorithm as a population
method where a local search operator is applied to each offspring
generated. It would be oversimplifying to state that memetic algorithms
are simply adding a local search operator to a population method.
There is a lot more behind MAs and this is testified by the success of the
many applications that have been published since the 90’s. In the case
of MAs, the local search operator is clearly a mean for intensification
(getting closer to the optimal solution) whereas the crossover operator,
initially designed in GAs for intensification, could also play a role as a
diversification operator."
"(i) The cognitive influence coefficient, 𝜑
, is defined in Equation (2). The cognitive influence coefficient is determined by a user’s intrinsic and social cognitive characteristics. We chose the personal character as 𝐷0
to describe users’ intrinsic cognitive characteristic, with 𝐷0∈ [0,1]
, where 0 means a user is extremely introverted and 1 means a user is extremely extroverted. Randomly generated values from 0
to 1
are used to represent varying personal characters. Here, 𝑗
is the in-degree neighbor of node 𝑖
, diff(𝑜𝑖,𝑜𝑗)
is the opinion difference between node 𝑖
and node 𝑗
, and 𝑘
is the number of in-degree neighbors of node 𝑖" -LOMDP: Maximizing Desired Opinions in Social Networks by Considering User Expression Intentions
The goal of the journal is to be an outlet for high quality theoretical and applied research on hybrid, knowledge-driven computational approaches that may be characterized under any of the following categories of memetics:
'Memetic Computing (Journal)'
abstract:
Type 1: General-purpose algorithms integrated with human-crafted heuristics that capture some form of prior domain knowledge; e.g., traditional memetic algorithms hybridizing evolutionary global search with a problem-specific local search. The journal welcomes investigations into various modes of meme transmission. Demonstrations of memetics in the context of deep neuroevolution, synergizing evolutionary search of neural architectures with lifetime learning of specific tasks or sets of tasks, are of significant interest.
Type 2: Algorithms with the ability to automatically select, adapt, and reuse the most appropriate heuristics from a diverse pool of available choices; e.g., learning a mapping between global search operators and multiple local search schemes, given an optimization problem at hand.
Type 3: Algorithms that autonomously learn with experience, adaptively reusing data and/or machine learning models drawn from related problems as prior knowledge in new target tasks of interest; examples include, but are not limited to, transfer learning and optimization, multi-task learning and optimization, or any other multi-X evolutionary learning and optimization methodologies.
Authors are encouraged to submit original research articles, including reviews and short communications, expanding the conceptual scope of memetics (e.g., to Type-X and beyond) and/or advancing the algorithmic state-of-the-art. Articles reporting novel real-world applications of memetics in areas including, but not limited to, multi-X evolutionary computation, neuroevolution, embodied cognition and intelligence of autonomous agents, continuous and discrete optimization, knowledge-guided machine learning, computationally expensive search problems, shall be considered for publication.
"To address this problem, we develop a theoretical
framework that models how users choose among multiple
sources of incoming information and affect the spreading of
memes on a directed social network, like Twitter [1–3]. Our
probabilistic model, in contrast to other studies [3–5,19,36]
that use intensive computational simulations to fit to data,
allows us to get analytical insights into the respective roles
of the network degree distribution, the memory-time
distribution of users, and the competition between memes
for the limited resource of user attention. The model is a
“null model” in the sense that it is analytically tractable, yet
realistic enough to be fitted to empirical data and to
reproduce some important characteristics of the data. We
show that fitting to time-dependent data requires a nontrivial memory-time distribution, which is not possible with
the toy model of Ref." -Effects of Network Structure, Competition and Memory Time
on Social Spreading Phenomena
