Minna no Data Sitesoil project

Why and how did MDS begin?

 


Why did Minnano Datasite begin?


In March, 2011, the accident of the Fukushima Daiichi nuclear plant began following the Eastern Japan earthquake.

Radioactive contamination was invisibly spread in eastern Japan and wider areas. In response to these circumstances,
a citizens’ movement for radioactivity measurement began and grew rapidly, run by people who volunteered to check
the contamination of food and other materials they produced and consumed. In some cases, volunteers shared
the cost of purchasing measurement instruments, while others asked for donations and applied for grants,
and ultimately many groups started to measure food, water, and soil in their areas.

In 2012, at a research exchange meeting on radioactive measurement conducted by the Takagi Fund
for Citizen Science for aid recipients and applicants, the Tokai No Nukes Network for
Future Generations Citizens' Radiation Measuring Center (C-lab) in Nagoya conceived the idea of producing a simple database
and systematized mapping of the measurement data.

C-lab initiated the project of constructing a full-scale database of independent radioactivity measurement laboratories.
In order to achieve this objective, C-lab and Takagi Fund consulted with Kodomomirai Radiation Measurement Station
for Children and Future (Kodomira RMS) which was managing the network of independent radioactivity measurement
laboratories in Japan.

These groups started to work together to establish and launch a joint food radioactivity measurement data site.


How to Overcome Data Integration Difficulties 

One of the problems was how to integrate the gathered data. 
As each radioactivity measurement lab used different measurement devices, the display format also varied from device to device. 
Therefore, it was very difficult to determine how to configure the different data, how to migrate them into what kind of database system, 
and also, how to output them with what kind of format. 

In addition, there was another problem on how to classify many ingredients and types of food that we are measuring. 
In order to classify foodstuffs, the MAFF, Ministry of Agriculture, Forestry and Fisheries of Japan, recommends using biological taxonomy, 
such as “Order” and “Class”, which ordinary people are not familiar with. 
That classification may be an obstacle for users to reach the data when they don’t know the scientific name.
Therefore, we have established and applied our own classification rules. 

Currently, "Minna no Data Site" offers food classification with three levels of categorized hierarchies. 
They are "upper item"(e.g. grain, potatoes), "middle item"(e.g. processed rice foods, processed potato food), 
and "lower item" (e.g. non-glutinous rice, sweet potato). 
Users can search items in any level. 

Furthermore, the Japanese language has three set of characters, which are hiragana, katakana, and kanji. 
Even if it is the same food, such as a “cucumber”, there are multiple notations in Japanese like “きゅうり”, ”胡瓜”, ”キュウリ”. 
In order to cope with this complexity, staff members of Minna no Data Site have clarified and registered all of the notations 
for each food type one by one. This great effort made the database system able to recognize and pickup different type of notation data. 

Another difficulty at the next stage was how to migrate the data already measured by each measurement lab into the database system. 
We asked the measurement labs which have relatively few measured data to input them into the system by hand. On the other hand,
for the measurement labs which have tremendous amount of data, we first extracted all of the data from each measurement device with a 
CSV format and then retrieved the data into the system after data cleansing. When the data migration failed, we checked each data one by one 
and detected and modified incomplete, inaccurate or irrelevant parts of the data manually. 


Website Release 

Thus, we advanced the final preparation towards release of the website. 
Since not all of the measurement labs have IT experts, our staff members who have good IT knowledge respectively 
and indefatigably supported labs to achieve migration tasks. It was a patient and time-consuming effort but it was realized through 
great cooperation beyond the barriers between measurement labs. 

When the migrated data almost reached 10,000 items, a plan for the full-scale release of the website was also drawn up. 
Around the same time, many measurement labs which had never been contacted before, have one after another applied 
to participate in our organization. So we felt a sense of conviction that a nationwide network like the current environment will be realized. 
However, as we moved forward one step further, we were also faced with the next problem. 


Difficulty in Disclosing Radioactive Contamination Data 

In the Kanto and Tokai areas, many food samples were offered mainly from housewives and mothers living in urban areas 
and paying attention to food safety. Since those samples covered a wide variety of food, many different kinds of data were accumulated. 

On the other hand, the amount of gathered data in the Tohoku area, which is the most crucial information, was not enough. 
Therefore, we wondered about the current environment in Tohoku and considered what we could do to help change this situation. 

There are several reasons why going public regarding radioactive contamination data for food products produced in the Tohoku area 
makes slow progress. One of the reasons is that there are many independent radioactivity measurement labs in Tohoku, 
that are not participating in Minna no Data Site. 

Some of those labs are membership societies which are run by local people and data is only open to the members who have paid 
an annual membership fee. They set management rules so that their measurement labs can be sustainably operated. 

In addition, some of measurement labs that are operated in a fairly small area such as a village do not disclose the data 
at all to protect local farmers or producers. This is because the producer might be identified if the food is a special local product. 
Those labs are concerned that even only one piece of data might widely spread and create reputational risk. 
Therefore, they decided to not disclose the data to protect local farmers/producers who kindly offered samples. 
This is what we learned about the current situation in Tohoku. 

Recognizing their sufferance special to contaminated areas is of some help to understand such problems as "eat or not eat" 
and "stay or evacuate" that are still controversial. This also includes some basic dilemmas such as "whether only one right answer truly exists"
and "whether non-local people could give directions to the local people who are suffering".