October 5, 2016
Thank you again for all the contributions to the Agricultural Open Data Package at the GODAN Summit 2016 ! We have processed your input and would like to share some of the results.
During the morning session on 15th September, we explained the process and the key considerations used for developing the Agricultural Open Data Package. We also consulted the audience on key dataset categories for agriculture. Using a short quiz we invited our attendees to select 3-7 priority datasets for the Agriculture Open Data Package. The top 5 data categories selected by a total of 34 respondents were Geospatial Base data (70%), Meteorological Data (59%), Research Data (55%), Market and Price Data (52%) and Agronomic Data (41%). Compared to our online consultation in June , we see similar priorities with a slightly different ordering. In the figure below combined both sets of results to get to a total of 63 respondents.
In the afternoon workshop we dug deeper with the objective of pinpointing the key datasets for the agricultural sector. The session consisted of 3 rounds.
- In the first round, participants were provided with a number of pre-selected data categories and examples of data sets. The participants were asked to value the importance of the datasets describing concrete use cases for the different end users in the agricultural sector.
- In the second round, we asked participants to discuss how easy or difficult it would be for governments to publish these different datasets, considering difficulty to collect, processing, sensitivities such as privacy implications
- In the third round, participants were asked to rank the data sets in order of priority based on the potential impact (the use) and the ease of publication of the data.
The evaluation was very challenging to complete because of the many datasets and their complexity. In general there was a feeling that most of the datasets (over 50) are very important for agriculture, confirming the challenge of getting to a shortlist of key datasets for the agriculture open data package.
It was also observed that the selection of key datasets depend on the perspective taken. As a result, we will develop a number of different policy perspectives to focus on, for example, optimizing agricultural practice or supporting agricultural finance data.
We also concluded that publishing data is often not the most challenging issue, instead we see more difficulties arise when attempting to get the data to a point of high quality and accuracy. The example of validating land use patterns or clarifying traditional land ownership rights demonstrated the enormous challenges we face.
During the workshop many standards and standard developing institutions have been shared as an addition to the package. With all of this in mind, we move onto IODC to further develop the Agriculture Open Data Package with the wider open data community focussing on implementation strategies.
Photo credit: Perry Bindelglass. CC BY-NC
This post originally appeared on the GODAN website.