2017-12-01ãã1ã¶æéã®è¨äºä¸è¦§
æ¥ã æ¥ããæ å ±ã¯å¤ãããä¸ã«ãã¾ã å¢å ãã¦ããããã®ãããªæ å ±éå¤(information overload)ã®åé¡ã«å¯¾ãã¦ãprecisionã¨recallã¨ããï¼ã¤ã®æ¦å¿µãåèã«æ¦ç¥ãç«ã¦ã¦ã¿ããã (ç»åï¼è±Wikipediaã®"Precision and Recall"ãã転è¼) Precisionã¨recallã¯ãâ¦
èªåã®ä¸ã«ã¯ç°ãªãï¼ã¤ã®ä¸»ç¾©(principlesãè¡åã®æé)ããããé·æéã®åå¨ã確èªãããããªãã§ãããããææã ãæã«å ¥ãã°ããã(outcome goal) ã¨ãã主義ã¨ããããå´©ããã¨ãã«ç¾ãããç©ã¿ä¸ãã¦ãããã¨ãåºãåã£ã¦éå½ã«å§ãããã(process goaâ¦
ãããåºç¤ã¨è¨ãããåå ã®å¦ç¿ãçµããã¨ã第äºè¨èªå¦ç¿è ã¯ãè±èªãå¦ã¶ã®ã§ã¯ãªããè±èªã§å¦ã¶ã®ã ããªã©ã®å©è¨ãåãããã¨ãå¤ãããã®ä¸æãèå³æ·±ãã®ã¯ãèªç¶è¨èªã«ããã¦æ¥é°è ã¨æãããã¡ãªå©è©(particles)ã主役ã¨ãªã£ã¦ãããããå¦ç¿è ãåæ°â¦