PATRICIA is a highly efficient algorithm for storing, indexing, and retrieving information in a large file. It is designed to be economical in terms of index space and reindexing time, and it does not require rearranging text or index when new material is added. The algorithm is flexible in the variety of keys it can respond to and is particularly effective for binary-coded libraries. PATRICIA retrieves information based on user-provided keys, with computational effort that is linear in the length of the key and the number of its occurrences, and independent of the size of the library. It has been implemented in various FORTRAN programs for the CDC-3600 and applied to several large information-retrieval problems. The algorithm constructs an index for a binary-coded library, allowing retrieval of targets based on keys. It modifies the index to reflect changes in the library, using minimal memory and computation. The index includes only numbers that designate locations in the text or index, and the computation required to find a key's occurrences is linear in the key's length and the number of its occurrences. Adding a new item involves copying the new item into the text, determining how much of it is already present, and updating the index with minimal effort. PATRICIA's design is based on the concept of chains and twins, which help in efficiently managing the index. The algorithm uses a system of coordinates to uniquely identify L-phrases, with chain coordinates and length coordinates. The index is structured to minimize storage and computational requirements, with the TWIN table being accessed most frequently. The algorithm's efficiency is further enhanced by the use of binary numbering for ends and branches, which simplifies storage and retrieval. Key components of PATRICIA include the START, STOP, END, L-PHRASE, BRANCH, TWIN, and CHAIN, which are used to organize and retrieve information. The algorithm's performance is optimized through the use of FINDONE and FINDALL, which efficiently locate and retrieve information based on keys. These algorithms ensure that the index remains efficient and adaptable, even as the library grows. The algorithm's design allows for minimal storage and computational overhead, making it suitable for large-scale information retrieval systems.PATRICIA is a highly efficient algorithm for storing, indexing, and retrieving information in a large file. It is designed to be economical in terms of index space and reindexing time, and it does not require rearranging text or index when new material is added. The algorithm is flexible in the variety of keys it can respond to and is particularly effective for binary-coded libraries. PATRICIA retrieves information based on user-provided keys, with computational effort that is linear in the length of the key and the number of its occurrences, and independent of the size of the library. It has been implemented in various FORTRAN programs for the CDC-3600 and applied to several large information-retrieval problems. The algorithm constructs an index for a binary-coded library, allowing retrieval of targets based on keys. It modifies the index to reflect changes in the library, using minimal memory and computation. The index includes only numbers that designate locations in the text or index, and the computation required to find a key's occurrences is linear in the key's length and the number of its occurrences. Adding a new item involves copying the new item into the text, determining how much of it is already present, and updating the index with minimal effort. PATRICIA's design is based on the concept of chains and twins, which help in efficiently managing the index. The algorithm uses a system of coordinates to uniquely identify L-phrases, with chain coordinates and length coordinates. The index is structured to minimize storage and computational requirements, with the TWIN table being accessed most frequently. The algorithm's efficiency is further enhanced by the use of binary numbering for ends and branches, which simplifies storage and retrieval. Key components of PATRICIA include the START, STOP, END, L-PHRASE, BRANCH, TWIN, and CHAIN, which are used to organize and retrieve information. The algorithm's performance is optimized through the use of FINDONE and FINDALL, which efficiently locate and retrieve information based on keys. These algorithms ensure that the index remains efficient and adaptable, even as the library grows. The algorithm's design allows for minimal storage and computational overhead, making it suitable for large-scale information retrieval systems.