id
stringlengths
6
6
query_type
stringclasses
14 values
question
dict
paraphrased_question
sequence
query
dict
template_id
stringclasses
8 values
query_shape
stringclasses
7 values
query_class
stringclasses
5 values
auto_generated
bool
2 classes
number_of_patterns
int32
1
14
AQ0508
Factoid
{ "string": "What are the titles and IDs of research papers that include a benchmark for the DDI extraction 2013 corpus dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"DDI extraction 2013 corpus\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ2344
Factoid
{ "string": "List the code links in papers that use the A2C+CoEX model in any benchmark?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"A2C+CoEX\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ2349
Factoid
{ "string": "Can you provide links to code used in papers that benchmark the Orthogonalized Soft VSM model?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Orthogonalized Soft VSM\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0169
Factoid
{ "string": "Can you list the models that have been evaluated on the SentEval dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"SentEval\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0119
Factoid
{ "string": "Can you list the models that have been evaluated on the COPA dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"COPA\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0179
Factoid
{ "string": "Could you provide a list of models that have been tested on the MPQA benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"MPQA\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1912
Factoid
{ "string": "Where can I find code references in papers that have used the Memory Compressed model for benchmarking purposes?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Memory Compressed\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1857
Factoid
{ "string": "What are the most commonly used benchmark datasets for the Image Generation research field?" }
[]
{ "sparql": "SELECT DISTINCT ?dataset ?dataset_lbl\nWHERE {\n ?problem a orkgc:Problem;\n rdfs:label ?problem_lbl. \n FILTER (str(?problem_lbl) = \"Image Generation\")\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n}" }
T06
Tree
WHICH-WHAT
true
5
AQ1751
Factoid
{ "string": "What is the name of the top performing model in terms of Score score when benchmarked on the Atari 2600 Defender dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Score\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Defender\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ1222
non-factoid
{ "string": "What is the top benchmark score and its metric on the Walker, walk (DMControl100k) dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Walker, walk (DMControl100k)\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ0891
Factoid
{ "string": "List the metrics that are used to evaluate models on the Habitat 2020 Point Nav test-std benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Habitat 2020 Point Nav test-std\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ1488
Factoid
{ "string": "What is the name of the top performing model in terms of Number of params score when benchmarked on the One Billion Word dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Number of params\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"One Billion Word\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ1090
non-factoid
{ "string": "What is the top benchmark score and its metric on the ChemProt dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ChemProt\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ0004
Factoid
{ "string": "Can you list the models that have been evaluated on the AI-KG dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"AI-KG\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1861
Factoid
{ "string": "What are the most commonly used benchmark datasets for the Multi-Task Learning research field?" }
[]
{ "sparql": "SELECT DISTINCT ?dataset ?dataset_lbl\nWHERE {\n ?problem a orkgc:Problem;\n rdfs:label ?problem_lbl. \n FILTER (str(?problem_lbl) = \"Multi-Task Learning\")\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n}" }
T06
Tree
WHICH-WHAT
true
5
AQ0332
Factoid
{ "string": "What models are being evaluated on the ModelNet40 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ModelNet40\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ2375
Factoid
{ "string": "Can you provide links to code used in papers that benchmark the VGG8B + LocalLearning + CO model?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"VGG8B + LocalLearning + CO\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1509
Factoid
{ "string": "Which model has achieved the highest Matched score on the MultiNLI benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Matched\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"MultiNLI\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ1837
Factoid
{ "string": "Provide a list of benchmarked datasets related to the Information Extraction research area?" }
[]
{ "sparql": "SELECT DISTINCT ?dataset ?dataset_lbl\nWHERE {\n ?problem a orkgc:Problem;\n rdfs:label ?problem_lbl. \n FILTER (str(?problem_lbl) = \"Information Extraction\")\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n}" }
T06
Tree
WHICH-WHAT
true
5
AQ0160
Factoid
{ "string": "Can you list the models that have been evaluated on the BIOSSES dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"BIOSSES\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0010
Factoid
{ "string": "What models are being evaluated on the TSE-NER dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"TSE-NER\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0268
Factoid
{ "string": "Can you list the models that have been evaluated on the Atari 2600 Asterix dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Asterix\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1294
non-factoid
{ "string": "What is the top benchmark score and its metric on the Atari 2600 Kangaroo dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Kangaroo\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ2007
Factoid
{ "string": "Provide a list of papers that have utilized the Ours: cross-sentence ALB model and include the links to their code?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Ours: cross-sentence ALB\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0965
Factoid
{ "string": "What are the metrics of evaluation over the Atari 2600 Road Runner dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Road Runner\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
HQ0040
Factoid
{ "string": "Which countries are considered in the papers about geopolitics?" }
[ "About which countries are there research papers on geopolitcs?" ]
{ "sparql": "SELECT DISTINCT ?location\nWHERE {\n ?_ orkgp:compareContribution [\n orkgp:P32 [\n rdfs:label ?label\n ];\n orkgp:P5049 ?location\n ]\n FILTER(REGEX(STR(?label), \"geopoli?tics\"))\n}" }
null
tree
WHAT-WHEN
false
4
AQ2448
Factoid
{ "string": "Can you provide links to code used in papers that benchmark the ViT-B/16 model?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"ViT-B/16\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1893
Factoid
{ "string": "List the datasets benchmarked under the relation extraction research problem?" }
[]
{ "sparql": "SELECT DISTINCT ?dataset ?dataset_lbl\nWHERE {\n ?problem a orkgc:Problem;\n rdfs:label ?problem_lbl. \n FILTER (str(?problem_lbl) = \"relation extraction\")\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n}" }
T06
Tree
WHICH-WHAT
true
5
AQ0057
Factoid
{ "string": "What are the models that have been benchmarked on the CoNLL04 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"CoNLL04\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1318
non-factoid
{ "string": "Can you provide the highest benchmark result, including the metric and score, for the Atari 2600 Gravitar dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Gravitar\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ1459
Factoid
{ "string": "What is the best performing model benchmarking the Natural Questions dataset in terms of F1 (Long) metric?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"F1 (Long)\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Natural Questions\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ1271
non-factoid
{ "string": "What is the top benchmark score and its metric on the Atari 2600 Pong dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Pong\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ0367
Factoid
{ "string": "Give me a list of research papers along with their titles and IDs, that have performed benchmarks on the Dataset mentions in Social Sciences dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Dataset mentions in Social Sciences\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ1693
Factoid
{ "string": "Indicate the model that performed best in terms of Score metric on the Atari 2600 Double Dunk benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Score\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Double Dunk\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0982
Factoid
{ "string": "List the metrics that are used to evaluate models on the Ohsumed benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Ohsumed\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ0801
Factoid
{ "string": "What evaluation metrics are commonly used when benchmarking models on the CommonsenseQA dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"CommonsenseQA\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ1190
non-factoid
{ "string": "Can you provide the highest benchmark result, including the metric and score, for the ImageNet dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ImageNet\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ0738
Factoid
{ "string": "What evaluation metrics are commonly used when benchmarking models on the TACRED dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"TACRED\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ1875
Factoid
{ "string": "List the datasets benchmarked under the Sentiment Analysis research problem?" }
[]
{ "sparql": "SELECT DISTINCT ?dataset ?dataset_lbl\nWHERE {\n ?problem a orkgc:Problem;\n rdfs:label ?problem_lbl. \n FILTER (str(?problem_lbl) = \"Sentiment Analysis\")\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n}" }
T06
Tree
WHICH-WHAT
true
5
AQ0184
Factoid
{ "string": "What models are being evaluated on the BC2GM dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"BC2GM\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1834
Factoid
{ "string": "List the datasets benchmarked under the Automated Reinforcement Learning (AutoRL) research problem?" }
[]
{ "sparql": "SELECT DISTINCT ?dataset ?dataset_lbl\nWHERE {\n ?problem a orkgc:Problem;\n rdfs:label ?problem_lbl. \n FILTER (str(?problem_lbl) = \"Automated Reinforcement Learning (AutoRL)\")\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n}" }
T06
Tree
WHICH-WHAT
true
5
AQ1053
non-factoid
{ "string": "What is the top benchmark result (metric and value) over the dataset Dataset mentions in Social Sciences?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Dataset mentions in Social Sciences\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ2043
Factoid
{ "string": "Where can I find code references in papers that have used the Image Transformer model for benchmarking purposes?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Image Transformer\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1682
Factoid
{ "string": "Indicate the model that performed best in terms of ROUGE-L metric on the GigaWord benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"ROUGE-L\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"GigaWord\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ1796
Factoid
{ "string": "What is the name of the top performing model in terms of Accuracy score when benchmarked on the Recipe dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Accuracy\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Recipe\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ2303
Factoid
{ "string": "Where can I find code references in papers that have used the DARQN soft model for benchmarking purposes?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"DARQN soft\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0556
Factoid
{ "string": "Give me a list of research papers along with their titles and IDs, that have performed benchmarks on the Food-101 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Food-101\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ2244
Factoid
{ "string": "Provide a list of papers that have utilized the Linear SVM model and include the links to their code?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Linear SVM\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0477
Factoid
{ "string": "What are the titles and IDs of research papers that include a benchmark for the ARC (Challenge) dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ARC (Challenge)\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ0621
Factoid
{ "string": "List the title and ID of research papers that contain a benchmark over the Atari 2600 Battle Zone dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Battle Zone\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ1446
Factoid
{ "string": "What is the best performing model benchmarking the HMDB51 dataset in terms of Top-1 Accuracy metric?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Top-1 Accuracy\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"HMDB51\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0832
Factoid
{ "string": "List the metrics that are used to evaluate models on the SNLI benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"SNLI\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ1814
Factoid
{ "string": "Which model has achieved the highest Top 5 Accuracy score on the ObjectNet benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Top 5 Accuracy\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ObjectNet\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ1749
Factoid
{ "string": "Indicate the model that performed best in terms of Score metric on the Atari 2600 Pitfall! benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Score\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Pitfall!\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0770
Factoid
{ "string": "List the metrics that are used to evaluate models on the IWSLT2015 German-English benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"IWSLT2015 German-English\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ0772
Factoid
{ "string": "What are the metrics of evaluation over the WMT2016 Russian-English dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"WMT2016 Russian-English\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ1963
Factoid
{ "string": "Can you provide links to code used in papers that benchmark the LibLinear model?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"LibLinear\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0183
Factoid
{ "string": "What models are being evaluated on the BC5CDR-chemical dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"BC5CDR-chemical\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0783
Factoid
{ "string": "What are the metrics of evaluation over the Multimodal PISA dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Multimodal PISA\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ2381
Factoid
{ "string": "Provide a list of papers that have utilized the EffNet-L2 (SAM) model and include the links to their code?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"EffNet-L2 (SAM)\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1959
Factoid
{ "string": "List the code links in papers that use the EneRex model in any benchmark?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"EneRex\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ2053
Factoid
{ "string": "Provide a list of papers that have utilized the AVID+CMA (Modified R2+1D-18 on Kinetics) model and include the links to their code?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"AVID+CMA (Modified R2+1D-18 on Kinetics)\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ2312
Factoid
{ "string": "Where can I find code references in papers that have used the DQN Best model for benchmarking purposes?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"DQN Best\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ2144
Factoid
{ "string": "Can you provide links to code used in papers that benchmark the 12-layer Character Transformer Model model?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"12-layer Character Transformer Model\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1123
non-factoid
{ "string": "What is the highest benchmark result achieved on the STL-10 dataset, including the metric and its value?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"STL-10\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ0212
Factoid
{ "string": "Can you list the models that have been evaluated on the CINIC-10 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"CINIC-10\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1700
Factoid
{ "string": "Which model has achieved the highest Score score on the Atari 2600 Berzerk benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Score\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Berzerk\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0229
Factoid
{ "string": "What models are being evaluated on the X-Sum dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"X-Sum\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ1343
non-factoid
{ "string": "What is the highest benchmark result achieved on the VTAB-1k dataset, including the metric and its value?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"VTAB-1k\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ1565
Factoid
{ "string": "What is the name of the top performing model in terms of F1 entity level score when benchmarked on the BC5CDR-chemical dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"F1 entity level\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"BC5CDR-chemical\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ2184
Factoid
{ "string": "Provide a list of papers that have utilized the Rfa-Gate-Gaussian-Stateful (Small) model and include the links to their code?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Rfa-Gate-Gaussian-Stateful (Small)\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1927
Factoid
{ "string": "Where can I find code references in papers that have used the Luna model for benchmarking purposes?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"Luna\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1533
Factoid
{ "string": "Which model has achieved the highest Bits per byte score on the The Pile benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Bits per byte\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"The Pile\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0375
Factoid
{ "string": "List the title and ID of research papers that contain a benchmark over the Annotated development corpus dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Annotated development corpus\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ0950
Factoid
{ "string": "List the metrics that are used to evaluate models on the Atari 2600 Bank Heist benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Bank Heist\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ0051
Factoid
{ "string": "Could you provide a list of models that have been tested on the WebNLG benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"WebNLG\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
HQ0013
Factoid/Superlative
{ "string": "What was the most common type of approach for summarization before 2002?" }
[ "Which was the most popular approach for summarization until 2002?" ]
{ "sparql": "SELECT ?approach ?approach_label\nWHERE {\n orkgr:R6948 orkgp:compareContribution ?cont.\n ?cont orkgp:P15 ?implementation.\n ?implementation orkgp:P5043 ?approach.\n ?approach rdfs:label ?approach_label.\n}\nORDER BY DESC(COUNT(?approach_label))\nLIMIT 1" }
null
chain
WHICH-WHAT
false
4
AQ1443
Factoid
{ "string": "Which model has achieved the highest Pre-Training Dataset score on the UCF101 benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Pre-Training Dataset\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"UCF101\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0744
Factoid
{ "string": "What evaluation metrics are commonly used when benchmarking models on the ACE 2004 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ACE 2004\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ2085
Factoid
{ "string": "List the code links in papers that use the FusionNet (single model) model in any benchmark?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"FusionNet (single model)\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0273
Factoid
{ "string": "What models are being evaluated on the Atari 2600 Tennis dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Tennis\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
HQ0061
Non-factoid
{ "string": "Which vegetables are utilized for betanin extraction?" }
[ "What vegetables are used in the process of betanin extraction?" ]
{ "sparql": "SELECT ?vegetables, ?vegetables_labels\nWHERE {\n orkgr:R75363 orkgp:compareContribution ?contrib.\n ?contrib orkgp:P35147 ?compounds.\n ?compounds rdfs:label ?compounds_labels.\n FILTER(REGEX(?compounds_labels, \"etanin\"))\n ?contrib orkgp:P35148 ?vegetables.\n ?vegetables rdfs:label ?vegetables_labels.\n}" }
null
tree
WHICH-WHAT
false
5
AQ1619
Factoid
{ "string": "What is the name of the top performing model in terms of Top 1 Accuracy score when benchmarked on the ImageNet dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?metric a orkgc:Metric;\n rdfs:label ?metric_lbl.\n FILTER (str(?metric_lbl) = \"Top 1 Accuracy\")\n {\n SELECT ?model ?model_lbl\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ImageNet\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value;\n orkgp:HAS_METRIC ?metric.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.\n }\n ORDER BY DESC(?value)\n LIMIT 1\n }\n}" }
T05
Tree
WHICH-WHAT
true
12
AQ0724
Factoid
{ "string": "What are the metrics of evaluation over the Car speed in Liuliqiao District, Beijing dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Car speed in Liuliqiao District, Beijing\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
HQ0086
Non-factoid
{ "string": "What are the research problems Vernier Effect is related to?" }
[ "What is the list of research problems related to Vernier effect?" ]
{ "sparql": "SELECT DISTINCT ?problems, ?problems_labels\nWHERE {\n ?papers rdf:type orkgc:Paper.\n ?papers rdfs:label ?papers_labels.\n FILTER(REGEX(?papers_labels, \"Vernier Effect\", \"i\"))\n ?papers orkgp:P31 ?contrib.\n ?contrib orkgp:P32 ?problems.\n ?problems rdfs:label ?problems_labels.\n}" }
null
tree
WHICH-WHAT
false
5
AQ2460
Factoid
{ "string": "Are there any research problems with benchmark datasets in the realm of Natural Language Processing research?" }
[]
{ "sparql": "SELECT DISTINCT ?problem ?problem_lbl\nWHERE {\n ?rf a orkgc:ResearchField;\n rdfs:label ?rf_label.\n FILTER (str(?rf_label) = \"Natural Language Processing\")\n ?paper orkgp:P30 ?rf;\n orkgp:P31 ?cont.\n ?cont orkgp:HAS_BENCHMARK ?benchmark;\n orkgp:P32 ?problem.\n ?problem rdfs:label ?problem_lbl.\n}" }
T08
Tree
WHICH-WHAT
true
5
AQ0008
Factoid
{ "string": "What models are being evaluated on the Scholarly entity usage detection dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Scholarly entity usage detection\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0418
Factoid
{ "string": "Give me a list of research papers along with their titles and IDs, that have performed benchmarks on the WMT2014 English-French dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"WMT2014 English-French\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ1987
Factoid
{ "string": "Where can I find code references in papers that have used the ETL-Span model for benchmarking purposes?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"ETL-Span\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ0252
Factoid
{ "string": "What are the models that have been benchmarked on the Atari 2600 Asteroids dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"Atari 2600 Asteroids\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0037
Factoid
{ "string": "Could you provide a list of models that have been tested on the smallNLP-KG benchmark dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"smallNLP-KG\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0302
Factoid
{ "string": "Can you list the models that have been evaluated on the HoC dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"HoC\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0803
Factoid
{ "string": "What evaluation metrics are commonly used when benchmarking models on the WebQuestions dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"WebQuestions\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ0062
Factoid
{ "string": "Can you list the models that have been evaluated on the SemEval-2010 Task 8 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"SemEval-2010 Task 8\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6
AQ0870
Factoid
{ "string": "Can you list the metrics used to evaluate models on the BC2GM dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"BC2GM\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n}" }
T03
Tree
WHICH-WHAT
true
6
AQ0401
Factoid
{ "string": "List the title and ID of research papers that contain a benchmark over the ACE 2004 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"ACE 2004\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ0465
Factoid
{ "string": "What are the titles and IDs of research papers that include a benchmark for the SQuAD1.1 dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?paper ?paper_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"SQuAD1.1\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?paper orkgp:P31 ?cont;\n rdfs:label ?paper_lbl.\n}" }
T02
Tree
WHICH-WHAT
true
5
AQ1973
Factoid
{ "string": "Can you provide links to code used in papers that benchmark the SciBERT + CNN model?" }
[]
{ "sparql": "SELECT DISTINCT ?code\nWHERE {\n ?model a orkgc:Model;\n rdfs:label ?model_lbl.\n FILTER (str(?model_lbl) = \"SciBERT + CNN\")\n ?benchmark orkgp:HAS_DATASET ?dataset.\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n ?cont orkgp:HAS_MODEL ?model;\n orkgp:HAS_SOURCE_CODE ?code.\n}" }
T07
Tree
WHICH-WHAT
true
4
AQ1057
non-factoid
{ "string": "What is the top benchmark result (metric and value) over the dataset CS-NER?" }
[]
{ "sparql": "SELECT DISTINCT ?metric ?metric_lbl (MAX(?value) AS ?score)\nWHERE {\n {\n SELECT ?metric ?metric_lbl ?value\n WHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"CS-NER\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?eval orkgp:HAS_VALUE ?value.\n OPTIONAL {?eval orkgp:HAS_METRIC ?metric.\n ?metric rdfs:label ?metric_lbl.}\n ?cont orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?cont orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n }\n ORDER BY DESC(?value)\n }\n}\nGROUP BY ?metric ?metric_lbl" }
T04
Tree
WHICH-WHAT
true
13
AQ0001
Factoid
{ "string": "What models are being evaluated on the SemEval-2018 Task 7 dataset dataset?" }
[]
{ "sparql": "SELECT DISTINCT ?model ?model_lbl\nWHERE {\n ?dataset a orkgc:Dataset;\n rdfs:label ?dataset_lbl.\n FILTER (str(?dataset_lbl) = \"SemEval-2018 Task 7 dataset\")\n ?benchmark orkgp:HAS_DATASET ?dataset;\n orkgp:HAS_EVALUATION ?eval.\n ?paper orkgp:HAS_BENCHMARK ?benchmark.\n OPTIONAL {?paper orkgp:HAS_MODEL ?model.\n ?model rdfs:label ?model_lbl.}\n}" }
T01
Tree
WHICH-WHAT
true
6

Dataset Card for SciQA

Dataset Summary

SciQA contains 2,565 SPARQL query - question pairs along with answers fetched from the open research knowledge graph (ORKG) via a Virtuoso SPARQL endpoint, it is a collection of both handcrafted and autogenerated questions and queries. The dataset is split into 70% training, 10% validation and 20% test examples.

Dataset Structure

Data Instances

An example of a question is given below:

{
    "id": "AQ2251",
    "query_type": "Factoid",
    "question": {
        "string": "Provide a list of papers that have utilized the Depth DDPPO model and include the links to their code?"
    },
    "paraphrased_question": [],
    "query": {
        "sparql": "SELECT DISTINCT ?code\nWHERE {\n  ?model    a  orkgc:Model;\n            rdfs:label    ?model_lbl.\n  FILTER (str(?model_lbl) = \"Depth DDPPO\")\n  ?benchmark      orkgp:HAS_DATASET        ?dataset.\n  ?cont           orkgp:HAS_BENCHMARK      ?benchmark.\n  ?cont           orkgp:HAS_MODEL          ?model;\n                  orkgp:HAS_SOURCE_CODE    ?code.\n}"
    },
    "template_id": "T07",
    "auto_generated": true,
    "query_shape": "Tree",
    "query_class": "WHICH-WHAT",
    "number_of_patterns": 4,
}

Data Fields

  • id: the id of the question
  • question: a string containing the question
  • paraphrased_question: a set of paraphrased versions of the question
  • query: a SPARQL query that answers the question
  • query_type: the type of the query
  • query_template: an optional template of the query
  • query_shape: a string indicating the shape of the query
  • query_class: a string indicating the class of the query
  • auto_generated: a boolean indicating whether the question is auto-generated or not
  • number_of_patterns: an integer number indicating the number of gtaph patterns in the query

Data Splits

The dataset is split into 70% training, 10% validation and 20% test questions.

Additional Information

Licensing Information

SciQA is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0).

Citation Information

@Article{SciQA2023,
  author={Auer, S{\"o}ren
  and Barone, Dante A. C.
  and Bartz, Cassiano
  and Cortes, Eduardo G.
  and Jaradeh, Mohamad Yaser
  and Karras, Oliver
  and Koubarakis, Manolis
  and Mouromtsev, Dmitry
  and Pliukhin, Dmitrii
  and Radyush, Daniil
  and Shilin, Ivan
  and Stocker, Markus
  and Tsalapati, Eleni},
  title={The SciQA Scientific Question Answering Benchmark for Scholarly Knowledge},
  journal={Scientific Reports},
  year={2023},
  month={May},
  day={04},
  volume={13},
  number={1},
  pages={7240},
  abstract={Knowledge graphs have gained increasing popularity in the last decade in science and technology. However, knowledge graphs are currently relatively simple to moderate semantic structures that are mainly a collection of factual statements. Question answering (QA) benchmarks and systems were so far mainly geared towards encyclopedic knowledge graphs such as DBpedia and Wikidata. We present SciQA a scientific QA benchmark for scholarly knowledge. The benchmark leverages the Open Research Knowledge Graph (ORKG) which includes almost 170,000 resources describing research contributions of almost 15,000 scholarly articles from 709 research fields. Following a bottom-up methodology, we first manually developed a set of 100 complex questions that can be answered using this knowledge graph. Furthermore, we devised eight question templates with which we automatically generated further 2465 questions, that can also be answered with the ORKG. The questions cover a range of research fields and question types and are translated into corresponding SPARQL queries over the ORKG. Based on two preliminary evaluations, we show that the resulting SciQA benchmark represents a challenging task for next-generation QA systems. This task is part of the open competitions at the 22nd International Semantic Web Conference 2023 as the Scholarly Question Answering over Linked Data (QALD) Challenge.},
  issn={2045-2322},
  doi={10.1038/s41598-023-33607-z},
  url={https://doi.org/10.1038/s41598-023-33607-z}
}

Contributions

Thanks to @YaserJaradeh for adding this dataset.

Downloads last month
125
Edit dataset card