def search_content_formula(mathML):
(query_sem_fea, query_struc_fea, query_cn_fea,
query_var_fea) = features_extraction(mathML)
query_ino_terms = ino_sem_terms(query_sem_fea) # sematic term in order
query_sort_terms = sort_sem_terms(
query_sem_fea) # sematic term in sorted order
related_formulas = formula_retrieval(query_sort_terms)
query_ino_terms = [term for term in chain.from_iterable(query_ino_terms)]
query_sort_terms = query_sort_terms[len(query_sort_terms) - 1]
N = Formula.objects.count()
IDF_values = compute_IDF_values(query_ino_terms, query_sort_terms,
query_struc_fea, related_formulas, N)
results, num_of_results = formulas_ranking(query_ino_terms,
query_sort_terms,
query_struc_fea, query_cn_fea,
query_var_fea, related_formulas,
IDF_values, N)
return results, num_of_results
def search_content_formula(mathML):
(query_sem_fea, query_struc_fea, query_cn_fea, query_var_fea) = features_extraction(mathML)
query_ino_terms = ino_sem_terms(query_sem_fea) # sematic term in order
query_sort_terms = sort_sem_terms(query_sem_fea) # sematic term in sorted order
related_formulas = formula_retrieval(query_sort_terms)
query_ino_terms = [term for term in chain.from_iterable(query_ino_terms)]
query_sort_terms = query_sort_terms[len(query_sort_terms)-1]
N = Formula.objects.count()
IDF_values = compute_IDF_values(query_ino_terms, query_sort_terms,
query_struc_fea, related_formulas, N)
results, num_of_results = formulas_ranking(query_ino_terms, query_sort_terms, query_struc_fea,
query_cn_fea, query_var_fea, related_formulas,
IDF_values, N)
return results, num_of_results
def extractFormulaFeature(query):
"Input: Formula Query"
"Output: Query vector"
"""
Step 1: Query to MathML
"""
math_obj = asciitomathml.asciitomathml.AsciiMathML()
math_obj.parse_string(query)
mathML = math_obj.to_xml_string()
mathML = mathML.replace("<math xmlns=\"http://www.w3.org/1998/Math/MathML\">","<math>")
mathML = mathML.replace("&","")
"""
Step 2: MathML to Formula object
"""
#Extract four types of formula_obj
formula_obj = Formula()
(sem_features, struc_features, const_features, var_features) = features_extraction(mathML)
# Generate index terms
inorder_sem_terms = ino_sem_terms(sem_features)
sorted_sem_terms = sort_sem_terms(sem_features)
#Insert into formulas table
formula_obj.inorder_term = inorder_sem_terms
formula_obj.sorted_term = sorted_sem_terms
formula_obj.structure_term = struc_features
formula_obj.constant_term = const_features
formula_obj.variable_term = var_features
formula_obj.status = 1
"""
Step 3: Extract feature
"""
featureAll = readFeature('formula')
formula = formula_obj
formula.structure = []
formula.semantic = []
formula.constant = []
formula.variable = []
#semantic
if formula.sorted_term!= '[]':
f_semantic_array = formula.sorted_term[0]
for line in f_semantic_array:
fa = line.split('$')
for f in fa:
f = semantic_rep(f)
if f != "":
formula.semantic.append(f)
#structure
if formula.structure_term!= '[]':
f_structure_array = formula.structure_term
#print f_structure_array
for f in f_structure_array:
f = struct_rep(f)
if f != "":
formula.structure.append(f)
#constant
if formula.constant_term!= '[]':
f_constant_array = formula.constant_term
#print f_constant_array
for f in f_constant_array:
f = const_rep(f)
if f != "":
formula.constant.append(f)
#variable
if formula.variable_term!= '[]':
f_variable_array = formula.variable_term
for f in f_variable_array:
f = var_rep(f)
if f != "":
formula.variable.append(f)
"""
Step 4: Build vector
"""
line = []
#print formula.semantic
#print formula.structure
#print formula.constant
#print formula.variable
for s in readFeature('semantic'):
line.append(min(1,formula.semantic.count(s)))
for s in readFeature('structure'):
line.append(min(1,formula.structure.count(s)))
for c in readFeature('constant'):
line.append(min(1,formula.constant.count(c)))
for v in readFeature('variable'):
line.append(min(1,formula.variable.count(v)))
return line
