THE CHALLENGE OF INFORMATION RETRIEVAL

Information retrieval is now fully recognized as an urgent and ever‐growing problem. Our increasing knowledge threatens to become inaccessible and hence unknowable. Two great conferences have alerted the scientific world to the danger and have attempted to particularize the problems. Many workers in different fields have made claims for ‘systems’ as solutions, others, more cautious, are testing systems, and some are even carrying out work dignified by the name ‘research’. All these activities seem painfully haphazard, and little progress has been made; there are no clear lines of advance for the future. Such theoretical approaches as have been made are very elementary and their validity is much disputed. Surveys of current work are now providing a valuable aid to co‐ordination of effort, and yet it seems that most investigators are sticking to their own particular hunches and gimmicks and are ignorant of, or ignoring, the work of others. These are harsh strictures, and need supporting evidence; even then, destructive criticism, by itself, is a disservice to progress, and some constructive effort is needed. I shall therefore attempt first a review of the present position, and then, however inadequately, put forward some suggestions. There is little else that can be done in our present state of groping in the darkness before the dawn.