Experiments on the compression of dictionary entries

Summary form only given. The authors study the compression of a list of items which are access keys to a dictionary or encyclopedia. Redundancy can be either inside the single word or between words (depending on the order). Attention is focused on those compression algorithms which can exploit mainly these kinds of redundancies. Three sequential methods, the classic Huffman code, a variant of the LZW algorithm, and the cascade application of differential technique plus Huffman encoding are analysed. A modification of the last-mentioned method makes it sequential in N-length chunks and reduces access delays for real-time applications