差别

这里会显示出您选择的修订版和当前版本之间的差别。

--- zh:notes:math_softwares [2026/08/24 10:06]
pzczxs [Obtain Embedding from a Vector Model]
+++ zh:notes:math_softwares [2026/10/23 12:40] (当前版本)
pzczxs [原始数据]
@@ 行 1: / 行 1: @@
 ====== Math Software Dataset Construction ======
 ===== Citation Information =====
-  -
+  -Shuo Xu, Lanzhi Li, Jiahao Li, Xin An, Jianhua Liu, and Xuegang Deng, 2026. Evaluation of Mathematical Software Impact based on Diversity Measures. (Under Review)
@@ 行 38: / 行 38: @@
 The //text-embedding-v4// model is employed to produce a 64-demensional embedding vector for each word.
 <code bash>
-> python .\extract_embedding.py --model text-embedding-v4 --dimension 64 --output text-embedding-v4-64.txt --resume .\math_software.word.vocab
+> python .\extract_embedding.py --model text-embedding-v4 --dimension 64 --batch-size 10 --output text-embedding-v4-64.txt --resume .\math_software.word.vocab
 </code>
 ==== Count One- and Two- Grams for PMI Calculation ====
@@ 行 104: / 行 104: @@
 ==== text-embedding-v4 ====
 <code bash>
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/10/math_software -K 10
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/10/math_software -K 10
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/20/math_software -K 20
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/20/math_software -K 20
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/30/math_software -K 30
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/30/math_software -K 30
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/40/math_software -K 40
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/40/math_software -K 40
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/50/math_software -K 50
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/50/math_software -K 50
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/60/math_software -K 60
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/60/math_software -K 60
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/70/math_software -K 70
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/70/math_software -K 70
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/80/math_software -K 80
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/80/math_software -K 80
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/90/math_software -K 90
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/90/math_software -K 90
-> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 256 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/100/math_software -K 100
+> java -Xmx20g -jar GaussianATCreditModel.jar -sf -D 64 -e resources/qwen3.7-embeddings.zip -n 50 -sc 7 -fb data/math_software/100/math_software -K 100
 </code>
 ===== Calculate Diversity Indicators =====
-The following three diversity indicators are calculated: (1) Rao-Stirling ([[https://www.jstor.org/stable/25050293|Rao, 1982]]; [[https://doi.org/10.1098/rsif.2007.0213|Stirling, 2007]]), DIV ([[https://doi.org/10.1016/j.joi.2019.03.016|Leydesdorff et al., 2019]]), and Diversity ([[https://doi.org/10.1007/s11192-022-04336-3|Mutz, 2022]]).
+The following three diversity indicators are calculated: (1) Rao-Stirling ([[https://www.jstor.org/stable/25050293|Rao, 1982]]; [[https://doi.org/10.1098/rsif.2007.0213|Stirling, 2007]]), DIV ([[https://doi.org/10.1016/j.joi.2019.03.016|Leydesdorff et al., 2019]]), and div<sub>e</sub>([[https://doi.org/10.1007/s11192-022-04336-3|Mutz, 2022]]).
+==== 原始数据 ====
+  *<color blue>00software_new.csv</color>：id（软件编号），swmath_id（swmath编号），name（软件名称），description（软件描述）
+  *<color blue>00article.csv</color>：id（论文编号），title（论文标题），abst（论文摘要），doi（doi号），publication_year（出版时间），type（论文类型），language_code（语言）
+  *<color blue>00software_article.csv</color>：software_id（软件编号），article_id（论文编号），flag（0：软件论文，1：论文）；注意：一个软件对应一篇或多篇论文
+  *<color blue>00category.csv：id（类别id）</color>，name（类别名称）
+  *<color blue>00article_category.csv</color>：article_id（论文编号），category_id（类别id）；注意：一篇论文对应一个或多个类别
+==== 构建software_article_category关系表 ====
+  *筛选条件：
+    *flag为1
+    *language_code为en
+    *type为以下任意一种：“Article”, “Article; Book Chapter”, “Article; Early Access”, “Article; Proceedings Paper”, “Letter”, “Proceedings Paper”, “Review” or “Review; Book Chapter”
+    *title和abst不能为空
+  *代码：<color blue>01构建software_article_category.py</color>
+  *结果：<color blue>01software_article_category.csv</color>
+==== 统计每个软件对应的无重复论文数量和类别数量 ====
+  *代码：<color blue>统计02software_num_articles_categories.py</color>
+  *结果：<color blue>02software_num_articles_categories.csv</color>

硕风徐徐

用户工具

站点工具

差别

页面工具