Compression of CAI text giles
Ronald G. Ragsdale, Norman F. Salt · ACM SIGCUE Outlook · 1976
This study investigates a method of compressing the computer representation of computer assisted instruction (CAI) text files .Traditionally , each character of CAI text would be represented by at least seven binar y digits (bits) .With efficient text compression techniques, the averag e number of bits per character, and hence costs, can be reduced .For this study, various text compression algorithms were examine d and a technique of representing English words by fourteen bits was selected .For best compression results, the technique requires that CAI text b e written with a restricted vocabulary of 1,024 words or less .A vocabular y was selected based on Ogden ' s international language, Basic English .Six samples of CAI text from various institutions were used for th e study .The texts were rewritten in such a way that most of the words cam e from the restricted vocabulary .Simulation of the text compression wa s done with both versions of the text .Since many words in the original text are in the restricted vocabulary , some compression is achieved with the original text .The compressed representations of the original texts were 85% of the traditional representations , Thus, instead of representing each character by seven bits, an average o f less than six bits per character would be required .Although the rewritten texts were slightly longer, compression was bette r because almost all words were from the restricted vocabulary .The compresse d representations of the rewritten texts were 78% of the traditional representations of the original texts and 71% of the traditional representation s of the rewritten texts (less than five bits per character) .Of the three percentages, the relevant one for text compression depend s on the acceptability of using a restricted vocabulary for CAI text .Thi s acceptability depends on how the use of the restricted vocabulary affect s authors and students . Compression of CAI Text File sTwo of the costs inhibiting the growth of Computer Assisted Instructio n (CAI) are those of data storage and data communication for the text files .The cost of data storage continues to decline and may not present a seriou s problem in the future .The cost of data transmission, where terminals ar e located some distance from the central computer, is not declining and assume s p .6