AZ ORSZÁGOS SZÉCHÉNYI KÖNYVTÁR ÉVKÖNYVE 1959. Budapest (1961)

IV. Könyvtár- és művelődéstörténeti tanulmányok - Summaires

historical background and in this connexion points to the important pioneer activity in this field by our countryman, Jenő Porzsolt. Porzsolt built the first photo-composing machine in 1894, which served as the basis for further experimentation and research throughout the world. Parallel with the activity in this domain by the Russian, V. A. Gassiev, the work of another Hungarian, Ödön Über, can be considered to have been of prime importance. In the second part of his study, the author describes the following three groups of photo-composing systems, adding important bibliographical information: a) conversion of traditional type-setting into photo^composing (Intertype Photosetter and "Monotype"); b) New photo-composing methods ("Lumitype" and "Linofilm"); c) Earlier types of semi-automatic photo-composing machines ("Hadego" and "Hohlux-Typofot"). The third part of the study deals with the importance of photo-composing systems and investigates future possibilities. It has become evident that electronics will find access into the printing trade to an ever increasing extent. The use of photo-electric cells in modern printing has become a significant factor, simplifying and speeding up printing procedures. It is certain that photo-composing will revolutionize the technique of text transfer, and undoubtedly electronics will contribute to the acceleration of this process. Mechanical information retrieval based on the collation principle G. OROSZ The paper discusses the application of the standard Hollerith-machines for information retrieval based on collation. Collation is a type of information retrieval which is applied for Uniterm coordinate indexing or for optical punched cards (Sichtlochkarte). As many Hollerith cards are prepared for each document as descriptors selected for characterizing the document. Three zones are delimited on the cards. The first zone (A) includes the document number, the second (B) the code number, symbolizing the descriptor. The remaining larger part (C) can serve as space for marking in the bibliographical data of the document. This card collection can be stored in three different ways: either in order of sequence (the cards are stored in groups according to the descriptor markings, and within the groups, arranged according to document numbers), or partly ordered only (cards stored in groups as mentioned above, but without order within the groups), finally, they can be stored at random. As to selection, the latter method is the one least determined by conditions and thus the most general case. Therefore, the process of selection was drawn up to apply to this case in first line. Outlines of the selection method: 1. the sorter searches the card collection according to the descriptor markings of the subject of the reference question; 2. the sorter orders the selected card groups in succession according to document numbers; 3. the collator collates the ordered groups. The general solution is applicable for proceeding in the first two cases (with partly and fully ordered card collections), where certain steps can be omitted. 26* 399