add src and data

README.md

@@ -1,20 +1,43 @@
- # HMS: A Hierarchical Solver with Dependency-Enhanced Understanding for Math Word Problem
-Source code for paper *HMS: A Hierarchical Solver with Dependency-Enhanced Understanding for Math Word Problem*.
-The source code for paper "Learning Relation-Enhanced Hierarchical Solver for Math Word Problems" is coming soon.
-
- ## Dependencies
-- python >= 3.6
-
-- stanfordcorenlp
-- torch
-
- ## Usage
-Preprocess dataset
-```bash
-python3 src/dataprocess/math23k.py
-```
-Train and test model
-```bash
-python3 src/main.py
-```
-For running arguments, please refer to [src/config.py](src/config.py).
+ # Learning Relation-Enhanced Hierarchical Solver for Math Word Problems
+Source code for paper *Learning Relation-Enhanced Hierarchical Solver for Math Word Problems*.
+
+ ## Dependencies
+- python >= 3.6
+
+- stanfordcorenlp
+- torch
+
+ ## Usage
+- Preprocess dataset
+```bash
+python3 src/dataprocess/math23k.py
+python3 src/dataprocess/similarity.py
+```
+- Train and test model
+```bash
+python3 src/rhms/main.py
+```
+For running arguments, please refer to [src/rhms/config.py](src/rhms/config.py).
+
+ ## Citation
+If you find our work helpful, please consider citing our paper.
+```
+@article{lin2023learning,
+  title={Learning Relation-Enhanced Hierarchical Solver for Math Word Problems},
+  author={Lin, Xin and Huang, Zhenya and Zhao, Hongke and Chen, Enhong and Liu, Qi and Lian, Defu and Li, Xin and Wang, Hao},
+  journal={IEEE Transactions on Neural Networks and Learning Systems},
+  year={2023},
+  publisher={IEEE}
+}
+```
+```
+@inproceedings{lin2021hms,
+  title={HMS: A Hierarchical Solver with Dependency-Enhanced Understanding for Math Word Problem},
+  author={Lin, Xin and Huang, Zhenya and Zhao, Hongke and Chen, Enhong and Liu, Qi and Wang, Hao and Wang, Shijin},
+  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
+  volume={35},
+  number={5},
+  pages={4232--4240},
+  year={2021}
+}
+```

src/dataprocess/equ_tools.py

@@ -1,105 +1,105 @@
-# -*- encoding:utf-8 -*-
-
-def infix_to_postfix(expression):
-    st = list()
-    res = list()
-    priority = {"+": 0, "-": 0, "*": 1, "/": 1, "^": 2}
-    for e in expression:
-        if e in ["(", "["]:
-            st.append(e)
-        elif e == ")":
-            c = st.pop()
-            while c != "(":
-                res.append(c)
-                c = st.pop()
-        elif e == "]":
-            c = st.pop()
-            while c != "[":
-                res.append(c)
-                c = st.pop()
-        elif e in priority:
-            while len(st) > 0 and st[-1] not in ["(", "["] and priority[e] <= priority[st[-1]]:
-                res.append(st.pop())
-            st.append(e)
-        else:
-            res.append(e)
-    while len(st) > 0:
-        res.append(st.pop())
-    return res
-
-def postfix_to_prefix(post_equ, check=False):
-    op_list = set(["+", "-", "*", "/", "^"])
-    stack = []
-    for elem in post_equ:
-        sub_stack = []
-        if elem not in op_list:
-            sub_stack.append(elem)
-            stack.append(sub_stack)
-        else:
-            if len(stack) >= 2:
-                opnds = reversed([stack.pop() for i in range(2)])
-                sub_stack.append(elem)
-                for opnd in opnds:
-                    sub_stack.extend(opnd)
-                stack.append(sub_stack)
-    if check and len(stack) != 1:
-        pre_equ = None
-    else:
-        pre_equ = stack.pop()
-    return pre_equ
-
-def post_solver(post_equ):
-    op_list = set(['+', '-', '/', '*', '^'])
-    status = True
-    stack = []
-    for elem in post_equ:
-        if elem in op_list:
-            if len(stack) >= 2:
-                op = elem
-                opnd2 = stack.pop()
-                opnd1 = stack.pop()
-                if op == '+':
-                    answer = opnd1 + opnd2
-                elif op == '-':
-                    answer = opnd1 - opnd2
-                elif op == '*':
-                    answer = opnd1 * opnd2
-                elif op == '/':
-                    answer = opnd1 / opnd2
-                elif op == '^':
-                    answer = opnd1 ** opnd2
-                else:
-                    status = False
-                    break
-                stack.append(answer)
-            else:
-                status = False
-                break
-        else:
-            elem = float(elem)
-            stack.append(elem)
-    if status and len(stack) == 1:
-        answer = stack.pop()
-    else:
-        answer = None
-        status = False
-    return status, answer
-
-def number_map(equ, num_list):
-    num_equ = []
-    for token in equ:
-        if "temp_" in token:
-            token = num_list[ord(token[-1]) - ord('a')]
-        elif token == "PI":
-            token = 3.14
-        num_equ.append(token)
-    return num_equ
-
-def eval_num_list(str_num_list):
-    num_list = list()
-    for item in str_num_list:
-        if item[-1] == "%":
-            num_list.append(float(item[:-1]) / 100)
-        else:
-            num_list.append(eval(item))
-    return num_list
+# -*- encoding:utf-8 -*-
+
+def infix_to_postfix(expression):
+    st = list()
+    res = list()
+    priority = {"+": 0, "-": 0, "*": 1, "/": 1, "^": 2}
+    for e in expression:
+        if e in ["(", "["]:
+            st.append(e)
+        elif e == ")":
+            c = st.pop()
+            while c != "(":
+                res.append(c)
+                c = st.pop()
+        elif e == "]":
+            c = st.pop()
+            while c != "[":
+                res.append(c)
+                c = st.pop()
+        elif e in priority:
+            while len(st) > 0 and st[-1] not in ["(", "["] and priority[e] <= priority[st[-1]]:
+                res.append(st.pop())
+            st.append(e)
+        else:
+            res.append(e)
+    while len(st) > 0:
+        res.append(st.pop())
+    return res
+
+def postfix_to_prefix(post_equ, check=False):
+    op_list = set(["+", "-", "*", "/", "^"])
+    stack = []
+    for elem in post_equ:
+        sub_stack = []
+        if elem not in op_list:
+            sub_stack.append(elem)
+            stack.append(sub_stack)
+        else:
+            if len(stack) >= 2:
+                opnds = reversed([stack.pop() for i in range(2)])
+                sub_stack.append(elem)
+                for opnd in opnds:
+                    sub_stack.extend(opnd)
+                stack.append(sub_stack)
+    if check and len(stack) != 1:
+        pre_equ = None
+    else:
+        pre_equ = stack.pop()
+    return pre_equ
+
+def post_solver(post_equ):
+    op_list = set(['+', '-', '/', '*', '^'])
+    status = True
+    stack = []
+    for elem in post_equ:
+        if elem in op_list:
+            if len(stack) >= 2:
+                op = elem
+                opnd2 = stack.pop()
+                opnd1 = stack.pop()
+                if op == '+':
+                    answer = opnd1 + opnd2
+                elif op == '-':
+                    answer = opnd1 - opnd2
+                elif op == '*':
+                    answer = opnd1 * opnd2
+                elif op == '/':
+                    answer = opnd1 / opnd2
+                elif op == '^':
+                    answer = opnd1 ** opnd2
+                else:
+                    status = False
+                    break
+                stack.append(answer)
+            else:
+                status = False
+                break
+        else:
+            elem = float(elem)
+            stack.append(elem)
+    if status and len(stack) == 1:
+        answer = stack.pop()
+    else:
+        answer = None
+        status = False
+    return status, answer
+
+def number_map(equ, num_list):
+    num_equ = []
+    for token in equ:
+        if "temp_" in token:
+            token = num_list[ord(token[-1]) - ord('a')]
+        elif token == "PI":
+            token = 3.14
+        num_equ.append(token)
+    return num_equ
+
+def eval_num_list(str_num_list):
+    num_list = list()
+    for item in str_num_list:
+        if item[-1] == "%":
+            num_list.append(float(item[:-1]) / 100)
+        else:
+            num_list.append(eval(item))
+    return num_list