It's good to think through (1) who you would like to engage as your core group of 'early adopters' or interested partners, (2) why they would want to be part of a local data community and (3) how you plan to engage them. For example, you may want to focus on students because you are on a university campus. But why would they be interested in participating in your community: to add skills to their CV? To improve people's lives? And is the best way to reach them through events or online resources - or some other way?
There are different types of activities you may want to consider running (or joining), or some combination of these. Two or more events may even be linked together into a challenge. See these challenge examples which include multiple events: Healthy Environment, Spatial Transformation or Affordable Medicine. Or you may want to run a campaign to counter misinformation on social media.
An easy, fun way to get started is to hold a data Easter Egg Hunt. In this event, people form small teams that compete against each other to answer a set of questions by searching for relevant data.
Have a look at this one from Open Data Day 2016. It includes questions and datasets that you could use or adapt for your own community.
Usually an open 1-2 day event where participants form small groups to work with data and to create something new, such as a data analysis, mobile app, infographic, artwork. These are normally unstructured with occasional check-in points to evaluate progress. You may even want to do a follow-on hackathon 2-3 weeks later to build on the work done in the first event. Examples:
Geekulcha runs hackathons around South Africa through its local student societies. You can contact them if you would like to setup a student society on your campus.
Unconferences are a really good way to get diverse individuals together to define and explore issues and ideas related to a specific social theme (e.g. drought). It's participant-driven, so there is no fixed agenda. But you need to have a good facilitator to manage the conversation. The unconference can also be used to identify topics that a hackathon should focus on.
This article gives a short overview of what is needed to run an unconference. This South African team runs 1-day (or longer) 'Open Space Technology' events around the country. You may also consider running a World Cafe event or hosting/ joining a more traditional conference! See Eventbrite or Meetup for events in your area.
Is a more focused version of a hackathon that aims to solve a specific problem by guiding a group of selected participants towards certain design goals.
These are often closed events related to a specific project. They typically involve some form of human-centred design or design thinking process.
A data training over 1-2 days is a good way of introducing people to data topics and in encouraging data practitioners to explore how data can be applied to specific community needs. Most data training should not be technical, anyone should be able to participate. Examples:
TrainUp collaborates with School of Data who post their own content on extracting, cleaning, analysing, mapping, presenting data - including on specific themes like budget data.
There are also more technical online courses that you (or your community) could do together. For example, Datacamp offers a number of free introductory courses on Python, R and related topics.
You may want to host regular 2-3 hour meetings with your community after work or on weekends.
These are useful because people can plan ahead if they know it happens. Although these are social events it is good to provide some structure to the evening (such as speakers or projects to work on) so that newcomers know where to get involved and participants are able to achieve something. Codebridge uses hackdash to record and track progress on ideas emerging from hackathons. To see what is already happening in your area visit Meetup or Eventbrite sites. Examples of data, tech and innovation hubs around South Africa - which usually have open events:
Students have long holiday breaks and this is a great time to get them involved in VacWork - if they are staying on campus or coming home to your town. You may want to setup a programme over 2 or more weeks where students come for training and participate in a hackathon, working on a specific problem. Examples:
A good way to start building a local community is by sharing open data, guides or content with your network. If it is web-based you can share with your network as an early draft and update as you source new content, rather than waiting until you have a final book to share as pdf or hardcopy. Examples:
In this example from NASA they explain how to track air pollution from space using satellite images and how to compare it to data from terrestrial monitoring stations.
We have been adding to a working catalogue of 300+ (mostly) public data portals and websites in South Africa that you browse and share with each other.
You will probably want to let people know that events are happening or resources are available. There are many different web-based tools for this. The real challenge is to ensure content and communication are relevant to your audience.
If you are looking to send a newsletter or setup a mailing list then Mailchimp is great, and free for less than 2000 subscribers. It also integrates with social media.
Open Data Durban regularly posts useful resources on their Twitter feed. They often repeat tweets about specific projects they are working on or local events, so people are more likely to see this info. They also find and post content from other local and international sources that is of interest to their community.
The Code for America Brigades have published useful resources for hosting a local data community.
If you would like to dig deeper into the strategy, the Sunlight Foundation has published "A Guide to Tactical Data Engagement" (TDE). This guide aims to assist (primarily government) local data champions "Facilitate the impactful use of open data by collaborating communities".
A CMRA report identifies a number of key open data needs from a community or civil society perspective. For example, the use of data should lead to dialogue and accountability.
The International Federation of Red Cross and Red Crescent Societies (IFRC) recently (late 2018) published a Data Playbook to improve data literacy, through 30-60 minute conversations.