Editorial

Citation

Wilson, H.C. (2001), "Editorial", Disaster Prevention and Management, Vol. 10 No. 2. https://doi.org/10.1108/dpm.2001.07310baa.001

Editorial

If there is one thing that disasters produce it is mountains of information. So much information that it is impossible for mere human beings to analyse properly. The volumes of material that can accumulate after a disaster together with the frequency of such events means that no one individual has the time to collate or analyse it in the manner it deserves.

As someone who is keen to research into the prediction and the post-events of all types of disasters I am frequently baulked by the sheer amount of reading that has to be done. Never mind the analysis, the amount of reading is enormous. This, I feel, deters many very capable researchers from tackling the problems in a systematic manner.

But help is at hand. A new system of computer programming known as Inductive Logic Programming (ILP), can help enormously. This is a relatively new field in programming and as yet there are not many people participating.

It works like this. The researcher programs the computer with as much expert information on the topic and if it is told something it can propose a theory based on its information base. The researcher does not have to pre-load all of the information the program will go "mining" for its own information from available resources. It will come back with its version of the outcome(s). This can then be further refined by the researcher and the cycle completed again. The theory proposed by the computer also comes with a rationale as to why this theory could be true. This is evaluated by the researcher who can accept or reject the theory.

The system is being used to investigate new potential chemical inhibitors for hypertension; other researchers are using a similar system to investigate the effects of gene alteration in yeast cells in which the computer researches the available information, designs the experiments, e-mails the information to a robotic yeast culturing system which performs the analysis and e-mails the results back to the computer. The computer then compares the results with its theoretical derivatives and throws out those that do not work, creates another set of experiments which it e-mails to the analysis machine, and so on, until the end point is reached. During this the researcher is kept informed of the outcomes.

Just imagine if we had such a system working on the prediction of earthquakes, volcanic eruptions, railway accidents, air disasters, flood prediction etc. Now, if we could isolate those theories that would allow timely intervention …

Henry C. Wilson