| 摘要 |
PURPOSE: To correctly predict end time while adjusting it corresponding to the progress of a processing even in the case of processing long sentences to be the object of a natural language processing altogether. CONSTITUTION: The entire document of the natural langugage of a processing object is inputted, character numbers for constituting respective sentences included in it are discriminated and the numbers of the sentences for the respective character numbers are set in a table 11. The natural language processing is performed for each sentence, the processing time is measured and result data are set at a position corresponding to the pertinent character number inside the table 11 at the time of completion. Every time the natural language processing of one sentence is ended, the processing time of a remaining sentence number for the respective character numbers is calculated by using the result data for the respective character numbers of the table 11, the processing time is calculated by using the result data for the character number without the result data and respective calculation values are totaled and outputted as the processing end time. |
