HXL 1.0 Beta Release: A simple standard for messy data

OCHA is excited to announce the 1.0 beta release of the Humanitarian Exchange Language (HXL) standard. It has been developed with the help of representatives from British Red Cross, the Humanitarian Innovation Fund, the International Organization for Migration, Save the Children, the Standby Task Force, the UK Department for International Development, the UN Refugee Agency, UNICEF, USAID, Ushahidi, the World Food Programme, and the World Bank.

 

How it works

Inspired by social-media hashtags, HXL helps humanitarian organisations add value to the data they are creating by improving automation and interoperability.  The principle of the standard is simple: we ask organisations to add a row of HXL hashtags to their spreadsheets, between the headers and the data, like this:

With the addition of these tags, it is possible for computers to recognise the type of data expected in a column and to merge data from different sources more accurately and efficiently. We hope this will decrease the time spent on data management and increase the insight from data analysis.

The HXL Tags

The HXL beta release includes 33 hashtags. They cover concepts common to many types of humanitarian data, such as geopolitical information (e.g., #region, #country, #adm1, #loc), population indicators (e.g., #affected, #inneed, #targeted), response data (e.g., #activity, #org, #sector, #impact, #capacity), and crisis description (e.g., #crisis, #cause).

To refine the tags, a data author can use attributes like +code, +children, +incamp, or +start; for example, #affected +f +children indicates the number of girls affected by a crisis in the area being described.

To support easy access in the field, the core standard is small enough to fit on a 4×6 inch (10×15 cm) card.  The HXL index card and design assets can be found here.





Front of the HXL postcard Back of the HXL postcard

Accompanying the standard’s beta release are public-domain Python and Javascript libraries for HXL processing (Java and PHP coming soon), as well as an early prototype online service for validating, filtering, transforming, and visualising HXL-tagged data on the fly.

Pilots and further development

We plan to further develop these tools and eventually, to integrate HXL support into the Humanitarian Data Exchange. We are also investigating building HXL support directly into data-gathering tools such as KoBo to add intelligence right at the point of collection.

The use of HXL hashtags in spreadsheets was tested during the West Africa Ebola Response and more recently with OCHA’s consolidated 3W data for the Nepal Earthquake.

We would like to get your feedback on this release. Do you think these are the right tags for humanitarian data? How do you see HXL helping you with your data work? Send feedback to hdx@un.org.