4 County OSM Digitization Liberia – Lesson Learned

Room: Track 2

Saturday, 10:45 UTC
Duration: 20 minutes (plus Q&A)

Back to schedule

Tri Selasa

The Government of Liberia is working on building an effective social protection by implementing a Liberia Household Social Registry (LHSR). The data collection for this Liberia Household Social Registry (LHSR) requires house-to-house visits. To ensure that the data intake questionnaire is administered to all households, updated map is needed since it’s last updated in 2008. HOT collaborated with OSM Liberia and iLab Liberia to complete the mapping and perform quality assurance in four counties in Liberia. The objective of this program is to update OpenStreetMap data to assist with a Social Registry data collection in four counties in Liberia: Bomi, Bong, Nimba and Maryland. Buildings and roads are the entities to map for this program. The mapping itself is done by local university students. After completed mapping all four counties, they perform data validation using JOSM. Then, the data is being validated again by OSM Liberia, iLab Liberia and HOT ID. Lessons learned from this digitization program are:

First, it’s really great to involve local mappers because they have local knowledge especially on mapping roads. The road’s classification in Africa is quite different, they even have their own OSM Wiki Page.
Second, it’s really useful to review the works of mappers while they’re mapping, so that we can point out if there is any mistake. We can also make a guideline out of this mistake we found, so that they don’t do it again in the future.
Last, it’s better to make new mappers get familiar with mapping first before jumping to data validation. The longer duration of mapping, the better validator they would become. Though it can be helped by providing a validation guideline, it’s better for them to understand the do’s and the dont’s, what’s right or wrong, what’s supposed to do and not to do when mapping basic objects like buildings and roads. To grasp this level of understanding requires some time of mapping. We still find warnings, errors, conflicts and unfixed geometry in the tasks validated by local university students. Yes, this world would be better if there are more validators, but it takes some mapping time to be a good validator, to understand what needs to be corrected and what’s not.