def test_find_closer_to_pivot(self):
input = "Havinga et al systematically changed n from 3.00 to 4.00 by synthesizing LaTl 3" \
" (n=3.00, T c =1.6 K), LaPb 3 (n=3.75, T c =4.1 K), and " \
"ThPb 3 (n=4.00, T c =5.6 K) and the solid solutions " \
"La (Tl 1−x Pb x ) 3 and (La 1−x Th x )Pb 3 ."
spans = [("LaTl 3", "<material>"), ("T c", "<tc>"), ("1.6 K", "<tcValue>"),
("LaPb 3", "<material>"), ("T c", "<tc>"), ("4.1 K", "<tcValue>"),
("ThPb 3", "<material>"), ("T c", "<tc>"), ("5.6 K", "<tcValue>"),
("La (Tl 1−x Pb x ) 3", "<material>"), ("(La 1−x Th x )Pb 3", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
target = VicinityResolutionResolver()
closer_entity = target.find_closer_to_pivot(materials[0], tc_values)
assert closer_entity != None
assert closer_entity.text == "1.6 K"
closer_entity = target.find_closer_to_pivot(materials[1], tc_values)
assert closer_entity != None
assert closer_entity.text == "1.6 K"
def test_vicinityResolution_5(self):
input = "In fact, apart from the very recent discovery of the occurrence of a superconducting " \
"phase at 200 K in sulfur hydride systems under ultrahigh pressures (up to 150 GPa) , " \
"the highest T c materials found up until now can be grouped into two families: the " \
"cuprates, with T c of up to 164 K [5] (in HgBa 2 Ca 2 Cu 3 O 9 at 30 GPa), and " \
"Fe-pnictides and -chalcogenides (FPC) with T c of up to 55 K [6]."
spans = [("200 K", "<tcValue>"), ("sulfur hydride", "<material>"), ("highest T c", "<tc>"),
("cuprates", "<class>"), ("T c", "<tc>"), ("up to 164 K", "<tcValue>"),
("HgBa 2 Ca 2 Cu 3 O 9", "<material>"), ("Fe-pnictides and -chalcogenides", "<class>"),
("T c", "<tc>"), ("up to 55 K", "<tcValue>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 2
assert str(relationships[0][0]) == "sulfur hydride"
assert str(relationships[0][1]) == "200 K"
assert str(relationships[1][0]) == "HgBa 2 Ca 2 Cu 3 O 9"
assert str(relationships[1][1]) == "up to 164 K"
def test_vicinityResolution_respectively_4(self):
input = "In this paper, we look at the Bi-based materials that have the chemical formula " \
"Bi 2 Sr 2 Ca n-1 Cu n O 2n+4 (BiSCCO) where n=1, 2, 3 gives the first three members of this " \
"class: Bi 2 Sr 2 CuO 6 , Bi 2 Sr 2 CaCu 2 O 8 and Bi 2 Sr 2 Ca 2 Cu 3 O 10 , with critical " \
"temperatures ( ) T c of 20 K, 85 K and 110 K respectively."
spans = [("Bi 2 Sr 2 Ca n-1 Cu n O 2n+4 (BiSCCO)", "<material>"),
("Bi 2 Sr 2 CuO 6", "<material>"), ("Bi 2 Sr 2 CaCu 2 O 8", "<material>"),
("Bi 2 Sr 2 Ca 2 Cu 3 O 10", "<material>"),
("T c", "<tc>"), ("20 K", "<tcValue>"), ("85 K", "<tcValue>"), ("110 K", "<tcValue>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 3
assert str(relationships[0][0]) == "Bi 2 Sr 2 CuO 6"
assert str(relationships[0][1]) == "20 K"
assert str(relationships[1][0]) == "Bi 2 Sr 2 CaCu 2 O 8"
assert str(relationships[1][1]) == "85 K"
assert str(relationships[2][0]) == "Bi 2 Sr 2 Ca 2 Cu 3 O 10"
assert str(relationships[2][1]) == "110 K"
def test_calculate_distances_2(self):
input = "Havinga et al systematically changed n from 3.00 to 4.00 by synthesizing LaTl 3. " \
"T c = 1.6 K is then found in LaPb 3."
spans = [("LaTl 3", "<material>"), ("T c", "<tc>"), ("1.6 K", "<tcValue>"),
("LaPb 3", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
target = VicinityResolutionResolver()
distances = target.calculate_distances(materials, tc_values, doc)
assert len(distances) == 2
assert distances[materials[0]][tc_values[0]] == 27.0
assert distances[materials[1]][tc_values[0]] == 23.5
def test_calculate_distances(self):
input = "Havinga et al systematically changed n from 3.00 to 4.00 by synthesizing LaTl 3" \
" (n=3.00, T c =1.6 K), LaPb 3 (n=3.75, T c =4.1 K), and " \
"ThPb 3 with T c =5.6 K and the solid solutions " \
"La (Tl 1−x Pb x ) 3 and (La 1−x Th x )Pb 3 ."
spans = [("LaTl 3", "<material>"), ("T c", "<tc>"), ("1.6 K", "<tcValue>"),
("LaPb 3", "<material>"), ("T c", "<tc>"), ("4.1 K", "<tcValue>"),
("ThPb 3", "<material>"), ("T c", "<tc>"), ("5.6 K", "<tcValue>"),
("La (Tl 1−x Pb x ) 3", "<material>"), ("(La 1−x Th x )Pb 3", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
target = VicinityResolutionResolver()
distances = target.calculate_distances(materials, tc_values, doc)
assert len(distances) == 5
assert distances[materials[0]][tc_values[0]] == 7.5
assert distances[materials[1]][tc_values[1]] == 7.5
assert distances[materials[2]][tc_values[2]] == 18
def test_vicinityResolution_respectively_2(self):
input = "The critical temperature T C = 4.7 K discovered for La 3 Ir 2 Ge 2 in this work is by about 1.2 K " \
"higher than that found for La 3 Rh 2 Ge 2 ."
spans = [("critical temperature", "<tc>"), ("T C", "<tc>"), ("4.7 K", "<tcValue>"),
("La 3 Ir 2 Ge 2", "<material>"),
("La 3 Rh 2 Ge 2", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 1
assert str(relationships[0][0]) == "La 3 Ir 2 Ge 2"
assert str(relationships[0][1]) == "4.7 K"
def test_vicinityResolution_respectively_missingEntities_1(self):
input = "Ba 1−x K x BiO 3−δ (BKBO) and BaPb 1−x Bi x O 3−δ (BPBO) are two such compounds that show T c 's " \
"of 30 K [1] and 13 K [2], respectively, with carrier concentrations as low as 2×10 " \
"21 cm −3 ."
spans = [("BaPb 1−x Bi x O 3−δ (BPBO)", "<material>"),
("T c", "<tc>"), ("30 K", "<tcValue>"), ("13 K", "<tcValue>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 1
assert str(relationships[0][0]) == "BaPb 1−x Bi x O 3−δ (BPBO)"
assert str(relationships[0][1]) == "30 K"
def test_vicinityResolution_3(self):
input = "Havinga et al systematically changed n from 3.00 to 4.00 by synthesizing LaTl 3" \
" (n=3.00, T c =1.6 K), LaPb 3 (n=3.75, T c =4.1 K), and " \
"ThPb 3 (n=4.00, T c =5.6 K) and the solid solutions " \
"La (Tl 1−x Pb x ) 3 and (La 1−x Th x )Pb 3 ."
spans = [("LaTl 3", "<material>"), ("T c", "<tc>"), ("1.6 K", "<tcValue>"),
("LaPb 3", "<material>"), ("T c", "<tc>"), ("4.1 K", "<tcValue>"),
("ThPb 3", "<material>"), ("T c", "<tc>"), ("5.6 K", "<tcValue>"),
("La (Tl 1−x Pb x ) 3", "<material>"), ("(La 1−x Th x )Pb 3", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 3
def test_vicinityResolution_missingOneEntity_1(self):
input = "Superconductivity has been discovered in metal diborides like MgB 2 (T c =39 K ), (Mo 0.96 Zr 0.04 ) " \
"0.85 B 2 (T c =8.2 K ), NbB 2 (T c =5.2 K [3]) and various other ternary borides ."
spans = [("MgB 2", "<material>"), ("T c", "<tc>"),
("(Mo 0.96 Zr 0.04 ) 0.85 B 2", "<material>"), ("T c", "<tc>"), ("8.2 K", "<tcValue>"),
("NbB 2", "<material>"), ("T c", "<tc>"), ("5.2 K", "<tcValue>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 2
assert str(relationships[0][0]) == "(Mo 0.96 Zr 0.04 ) 0.85 B 2"
assert str(relationships[0][1]) == "8.2 K"
assert str(relationships[1][0]) == "NbB 2"
assert str(relationships[1][1]) == "5.2 K"
def test_vicinityResolution_respectively_1(self):
input = "In the best cases a transition temperature of 38 K (zero resistance point), 25 K (zero " \
"resistance point) and 38 K (midpoint) were measured for CCO/STO, CCO/BCO and LSCO/LCO, " \
"respectively."
spans = [("38 K", "<tcValue>"), ("25 K", "<tcValue>"), ("38 K", "<tcValue>"),
("CCO/STO", "<material>"), ("CCO/BCO", "<material>"), ("LSCO/LCO", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 3
assert str(relationships[0][0]) == "CCO/STO"
assert str(relationships[0][1]) == "38 K"
assert str(relationships[1][0]) == "CCO/BCO"
assert str(relationships[1][1]) == "25 K"
assert str(relationships[2][0]) == "LSCO/LCO"
assert str(relationships[2][1]) == "38 K"
def test_vicinityResolution_4(self):
input = "The investigated MnSi films are in a thickness regime where the magnetic transition " \
"temperature T c assumes a thickness-independent enhanced value of 43 K as compared with " \
"that of bulk MnSi, where T c ≈ 29 K. A detailed refinement of the EXAFS data reveals that " \
"the Mn positions are unchanged, whereas the Si positions vary along the out-of-plane " \
"direction, alternating in orientation from unit cell to unit cell."
spans = [("MnSi films", "<material>"), ("T c", "<tc>"), ("43 K", "<tcValue>"),
("MnSi", "<material>"), ("T c", "<tc>"), ("29 K", "<tcValue>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
relationships = VicinityResolutionResolver().find_relationships(doc, materials, tc_values)
assert len(relationships) == 2
assert str(relationships[0][0]) == "MnSi films"
assert str(relationships[0][1]) == "43 K"
assert str(relationships[1][0]) == "MnSi"
assert str(relationships[1][1]) == "29 K"
def test_find_following_entity(self):
input = "Havinga et al systematically changed n from 3.00 to 4.00 by synthesizing LaTl 3" \
" (n=3.00, T c =1.6 K), LaPb 3 (n=3.75, T c =4.1 K), and " \
"ThPb 3 (n=4.00, T c =5.6 K) and the solid solutions " \
"La (Tl 1−x Pb x ) 3 and (La 1−x Th x )Pb 3 ."
spans = [("LaTl 3", "<material>"), ("T c", "<tc>"), ("1.6 K", "<tcValue>"),
("LaPb 3", "<material>"), ("T c", "<tc>"), ("4.1 K", "<tcValue>"),
("ThPb 3", "<material>"), ("T c", "<tc>"), ("5.6 K", "<tcValue>"),
("La (Tl 1−x Pb x ) 3", "<material>"), ("(La 1−x Th x )Pb 3", "<material>")]
doc = prepare_doc(input, spans)
materials = [entity for entity in filter(lambda w: w.ent_type_ in ['<material>'], doc)]
tc_values = [entity for entity in filter(lambda w: w.ent_type_ in ['<tcValue>'], doc)]
target = VicinityResolutionResolver()
following = target.find_following_entity(materials[1], tc_values)
assert following is not None
assert following.text == "4.1 K"
all_entities = list(filter(lambda w: w.ent_type_ is not "", doc))
following = target.find_following_entity(materials[2], all_entities)
assert following is not None
assert following.text == "T c"
following = target.find_following_entity(materials[4], all_entities)
assert following is None
following = target.find_following_entity(tc_values[0], tc_values, "<material>")
assert following is None
following = target.find_following_entity(tc_values[2], all_entities, "<material>")
assert following is not None
assert following.text == "La (Tl 1−x Pb x ) 3"
def process_sentence(self, words, spaces, spans):
text = ''.join([
words[i] + (' ' if spaces[i] else '')
for i in range(0, len(words))
])
# print("Processing: " + text)
doc = self.init_doc(words, spaces, spans)
extracted_entities = {}
# svg = displacy.render(doc, style="dep")
# filename = hashlib.sha224(b"Nobody inspects the spammish repetition").hexdigest()
# output_path = Path(str(filename) + ".svg")
# output_path.open("w", encoding="utf-8").write(svg)
# extracted_entities['tokens'] = words
### TC VALUES CLASSIFICATION
# self.markCriticalTemperature(doc)
### RELATIONSHIP EXTRACTION
extracted_entities['relationships'] = []
destination_entities = [
entity for entity in filter(
lambda w: w.ent_type_ in [self.destination], doc)
]
source_entities = [
entity for entity in filter(
lambda w: w.ent_type_ in [self.source] and w._.linkable is
True, doc)
]
## 1 simple approach (when only one temperature and one material)
resolver = SimpleResolutionResolver()
relationships = resolver.find_relationships(destination_entities,
source_entities)
if len(relationships) > 0:
extracted_entities['relationships'].extend(
RuleBasedLinker.collect_relationships(relationships, 'simple'))
# print(" Simple relationships " + str(extracted_entities['relationships']['simple']))
else:
## 2 vicinity matching
resolver = VicinityResolutionResolver()
relationships = resolver.find_relationships(
doc, destination_entities, source_entities)
if len(relationships) > 0:
extracted_entities['relationships'].extend(
RuleBasedLinker.collect_relationships(
relationships, 'vicinity'))
# print(" Vicinity relationships " + str(extracted_entities['relationships']['vicinity']))
# else:
## 3 dependency parsing matching
# resolver = DependencyParserResolutionResolver()
# relationships = resolver.find_relationships(destination_entities, source_entities)
# if len(relationships) > 0:
# extracted_entities['relationships'].extend(
# RuleBasedLinker.collect_relationships(relationships, 'dependency'))
# print(" Dep relationships " + str(extracted_entities['relationships']['dependency']))
converted_spans = [
span_to_dict(entity) for entity in filter(
lambda w: w.ent_type_ in entities_classes(), doc)
]
extracted_entities['spans'] = converted_spans
extracted_entities['text'] = text
return extracted_entities