introductiontoconditionalrandomfields-資料下載頁
2025-07-15 19:24本頁面
【正文】 :0) Org: 他們通過各種渠道引進技術、優(yōu)良品種， result: 他們通過各種渠道引進技術、優(yōu)良品種， (對 :18,錯 :0) Org: 建立了一批以熱帶、亞熱帶作物為主的商品生產基地。 result: 建立了一批以熱帶、亞熱帶作物為主的商品生產基地。 (對 :24,錯 :0) Org: 宣傳西瓜價格開放政策和市場需求量， result: 宣傳西瓜價格開放政策和市場需求量， (對 :17,錯 :0) Org: 畝產由原來的四百多公斤上升到五百五十多公斤， result: 畝產由原來的四百多公斤上升到五百五十多公斤， (對 :22,錯 :0) Some result Feature space ? Relations between feature space size and corpus size Fig. 1. Relation between the number of features and sentences. The top line describes the case of building feature spaces with an extern dictionary, and the bottom dashed line without the extern dictionary. Feature space ? Distribution of feature occurrence Fig. 2. Distribution of feature occurrence. From the top down, lines represent feature spaces built from 100, 60, 30 and 10 thousand sentences with dictionary for both of the two figures. Picture (a) illustrates the number of features which occur no more than some certain times, while picture (b) exactly some certain times. (a) (b) Feature space ? Distribution of feature weights Fig. 3. Distribution of feature weight. The model is trained on 40,000 sentences. Result ? Relation between testing accuracy and removing boundary Fig. 4. Relation between testing accuracy and removing boundary. The original model is trained on 40,000 sentences. The straight line indicates removing features that has a weight value less than boundary, while the dashed line greater than boundary. D E T A I L S A BO U T M O D E L S T ra i n i n g C o r p u s S e n t Ch a r Rg F n T i m e T rA T eA 0 . 5 K 8 , 0 1 8 1 8 0 0 , 8 5 0 0 0 . 0 3 . 3 7 1 . 0 0 0 0 0 . 6 7 2 5 1K 1 6 , 3 4 3 2 5 0 8 , 3 7 5 0 0 . 0 8 . 5 9 0 . 9 4 7 7 0. 7102 2K 3 1 , 6 8 6 3 5 4 3 , 8 3 2 0 0 . 2 2 . 2 7 0 . 9 1 1 3 0 . 7 3 8 4 5K 7 7 , 5 1 5 4 6 7 9 , 0 7 6 0 1 . 1 9 . 3 7 0 . 8 9 0 6 0 . 7 7 6 8 10K 1 5 0 , 8 7 4 5 1 , 2 9 7 , 4 4 9 0 3 . 2 3 . 0 1 0 . 9 1 4 2 0 . 8 1 5 2 20K 2 9 6 , 6 8 4 8 1 , 4 5 8 , 6 5 0 1 2 . 3 4 . 5 0 0 . 9 0 2 3 0 . 8 3 3 4 40K 5 9 7 , 7 9 2 5 3 , 7 7 3 , 6 6 7 5 0 . 3 6 . 5 3 0 . 9 5 4 5 0 . 8 7 9 2 Fig. 5. Training and Testing result with different corpuses. Result Thank you for your attention !
環(huán)評公示相關推薦
鄂ICP備17016276號-1