Letter to the Editor

Dear Sir, The review (Program, Oct. 1977) of the report by Groenewegen and MarsHall ‘optical character recognition: the use of OCR techniques in decentralized data collection for bibliographic information systems’ somewhat misses both the point and the outcome of the investigation. Indeed, one might be led tobelieve that the difficulties encountered outweighed the benefits of the technique in the environment under consideration. The reverse is true. The investigators found that ‘the results have shown the feasibility of using OCR as a relatively simple technique for the preparation of input to computerised bibliographic information systems on a decentralized basis by countries that do not have the facility for preparing magnetic tape input. The main consideration is that centres preparing input should have access to sound equipment and the correct materials. However, the experiment showed that even flawed input could be processed and corrected very quickly and at a fraction of what it would cost to punch the input centrally’ (page 45).