摘要 |
PROBLEM TO BE SOLVED: To provide a natural language processor which eliminates a trouble to designate a part of speech for each occurrence place, has enhanced accuracy of output and improves a user's understanding of the output and reduces the trouble of post-processing. SOLUTION: The user prepares learning requirements for being stored in a learning dictionary part 104 together with a term instructed to be learn by the user from character strings being the both or either of the preceding and subsequent contexts of the term instructed to be learn by the user in an inputted sentence, and analysis processing utilizing learning contents is performed only when the requirements are established in the preceding and subsequent contexts. |