All Relations Lead to Rome: Automated Knowledge Graph Creation and Question Generation
Paper • 2606.22645 • Published • 1
Datasets:
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: ArrowInvalid
Message: Failed to parse string: 'Syracuse' as a scalar of type double
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1837, in _prepare_split_single
writer.write_table(table)
~~~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
self._write_table(pa_table, writer_batch_size=writer_batch_size)
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
pa_table = table_cast(pa_table, self._schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2303, in cast_table_to_schema
cast_array_to_feature(
~~~~~~~~~~~~~~~~~~~~~^
table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
feature,
^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1852, in wrapper
return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
~~~~^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2143, in cast_array_to_feature
return array_cast(
array,
...<2 lines>...
allow_decimal_to_str=allow_decimal_to_str,
)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1854, in wrapper
return func(array, *args, **kwargs)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2006, in array_cast
return array.cast(pa_type)
~~~~~~~~~~^^^^^^^^^
File "pyarrow/array.pxi", line 1147, in pyarrow.lib.Array.cast
File "/usr/local/lib/python3.14/site-packages/pyarrow/compute.py", line 412, in cast
return call_function("cast", [arr], options, memory_pool)
File "pyarrow/_compute.pyx", line 604, in pyarrow._compute.call_function
File "pyarrow/_compute.pyx", line 399, in pyarrow._compute.Function.call
result = GetResultValue(
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: Failed to parse string: 'Syracuse' as a scalar of type double
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1683, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1869, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
id int64 | question_type string | relational string | complexity string | persona string | entity_one string | entity_two null | relation_one null | base_question string | question string | embedding string | fact string | fact_chunk null |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Factoid | Entity | Single | Novice | Kingdom of Portugal | null | null | When was the Kingdom of Portugal founded? | What year did Portugal become a kingdom? | [-0.0046953084, 0.0030868168, -0.021905478, 0.012211364, -0.011814388, -0.03486482, 0.009735837, 0.0044200304, -0.009028543, -0.050276924, -0.04245304, -0.008354605, 0.01086004, -2.3350445e-05, -0.013529181, -0.015609924, -0.0024929387, 0.0059210933, 0.013443547, 0.00067485915, 0.00044922563, -0.014109149, 0.0015565174... | This question cannot be answered using the data. | null |
2 | Factoid | Entity | Single | Intermediate | Kingdom of Portugal | null | null | When was the Kingdom of Portugal founded? | When was the Kingdom of Portugal officially established, particularly after the period of Roman rule in the Iberian Peninsula? | [-0.0058955005, 0.0107950205, -0.020101802, 0.0085368315, -0.0007397774, -0.016120275, 0.018207444, 0.002757085, -0.022283535, -0.06075795, -0.033096775, -0.01038497, 0.010409988, 0.0038017833, -0.010957845, -0.0069430117, 0.004181661, 0.015148021, 0.010253745, -0.00037885833, -0.0020800591, -0.01300452, 0.010559345, 0... | This question cannot be answered using the data. | null |
3 | Factoid | Entity | Single | Expert | Kingdom of Portugal | null | null | When was the Kingdom of Portugal founded? | In which year was the Kingdom of Portugal established? | [-0.0039745276, -0.0011941793, -0.024210112, 0.012206823, -0.011233957, -0.020063568, 0.021602329, 0.004912019, -0.010976977, -0.051074926, -0.046492588, -0.007804287, 0.0007531964, -0.0058737653, -0.026624879, -0.019935967, 0.005722721, 0.01745048, 0.010286801, -0.005002549, 0.002815906, -0.011661526, 0.0028758242, 0.... | This question cannot be answered using the data. | null |
4 | Open | Entity | Single | Novice | Kingdom of Portugal | null | null | How did Portugal’s Atlantic coastline and early maritime expertise combine with its monarchical goals to drive the country’s Age of Discoveries expansion? | How did Portugal’s coast and early sailing skills together with its king’s plans lead to more discoveries during the Age of Discoveries? | [0.012749839, 0.034339987, 0.007590214, -0.0029654675, 0.008061472, -0.022285491, 0.02461426, -0.0017532363, 0.0023315575, -0.046627976, -0.036387525, 0.0017512294, -0.006220528, 0.010103969, 0.005228518, -0.020674443, -0.011619915, -0.009934227, 0.0076376074, 0.008191762, 0.003693373, 0.0019402131, -0.0020491036, 0.01... | This question cannot be answered using the data. | null |
5 | Open | Entity | Single | Intermediate | Kingdom of Portugal | null | null | "How did Portugal’s Atlantic coastline and early maritime expertise combine with its monarchical g(...TRUNCATED) | "How did Portugal's Atlantic coastline and early maritime expertise align with its monarchs' goals t(...TRUNCATED) | "[0.00036574452, 0.045073923, -0.023655578, 0.012349296, 0.0032155665, -0.03095827, 0.028840868, -0.(...TRUNCATED) | This question cannot be answered using the data. | null |
6 | Open | Entity | Single | Expert | Kingdom of Portugal | null | null | "How did Portugal’s Atlantic coastline and early maritime expertise combine with its monarchical g(...TRUNCATED) | "In what ways did the Atlantic coastal configuration of Lusitania (modern Portugal) and its emerging(...TRUNCATED) | "[-0.026733348, 0.03899076, -0.013684089, 0.0128698535, 0.0035251311, -0.03436943, 0.019583885, -0.0(...TRUNCATED) | This question cannot be answered using the data. | null |
7 | Factoid | Entity | Single | Novice | Suleiman the Magnificent | null | null | In what year was Suleiman the Magnificent born? | When was Suleiman the Magnificent born? | "[0.0126858, 0.0020739986, 0.0048188176, -0.036265176, -0.044472277, -0.003548654, -0.0072197267, 0.(...TRUNCATED) | This question cannot be answered using the data. | null |
8 | Factoid | Entity | Single | Intermediate | Suleiman the Magnificent | null | null | In what year was Suleiman the Magnificent born? | What year was Suleiman the Magnificent born? | "[0.012305368, -0.0011286078, -0.000897224, -0.036721844, -0.04342861, -0.007135073, -0.009295457, 0(...TRUNCATED) | This question cannot be answered using the data. | null |
9 | Factoid | Entity | Single | Expert | Suleiman the Magnificent | null | null | In what year was Suleiman the Magnificent born? | In which year was Suleiman the Magnificent born? | "[0.014237097, -0.0013653581, 0.0006494733, -0.030524524, -0.04274246, -0.004688422, -0.0045651854, (...TRUNCATED) | This question cannot be answered using the data. | null |
10 | Open | Entity | Single | Novice | Suleiman the Magnificent | null | null | "How did Suleiman's legal codification efforts influence the administrative practices of the Ottoman(...TRUNCATED) | What impact did Suleiman's law‑making have on how the Ottoman Empire was run after he died? | "[0.020959746, -0.0027288015, -0.002411432, -0.0029571487, -0.017418882, -0.019552812, 0.012098263, (...TRUNCATED) | This question cannot be answered using the data. | null |
End of preview.
Description
All Relations Lead to Rome (ARLtR) includes a knowledge graph made in Neo4j, and multiple QA-pairs. The dataset is supplementary material for the paper "All Relations Lead to Rome: Automated Knowledge Graph Creation and Question Generation"
The knowledge graph is stored in a dump file, while the QA-pairs are stored in different csv files. Make sure to follow the setup guide to get started.
Setup
- Import the dump in Neo4j
- Host an instance of the Neo4j Database.
- Link with said instance in your application.
- Use Gemini-Embedding-2 to create embeddings for questions (or use your own embedding model).
- Then use these embeddings for retrieval, these are stored in the embedding attribute of the CHUNK and ENTITY nodes. Note: You can use any query method to retrieve from the dataset, to use in different applications (i.e. non-RAG).
QA-Pairs
- Load either or both of the csv files
- Use the 'question' column to retrieve a given question, and the 'fact' column to retrieve facts (i.e. answers).
- These can then be used for benchmarking, each question has a corresponding 'embedding' attribute.
Datastructure
The knowledge graph contains nodes of type ENTITY and CHUNK. With chunks having .text and .embedding attributes. For the different relations, refer to the paper. Alternatively, you can import the dataset and query all relation types.
Paper / Reference
You can read our paper on arXiv (https://arxiv.org/abs/2606.22645) [1] The paper introduces a general framework for constructing knowledge graphs suited for hybrid retrieval and semantic selections.
References
[1] Jansen op de Haar, M., Stähle, T., & Gatti, L. (2026). All Relations Lead to Rome: Automated Knowledge Graph Creation and Question Generation. arXiv [Cs.IR]. Retrieved from http://arxiv.org/abs/2606.22645
- Downloads last month
- 74