Add split miner implementation

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

README.rst

@@ -0,0 +1,137 @@
+===========
+split-miner
+===========
+
+
+Description
+===========
+
+**split-miner** is a Python implementation of the Split Miner algorithm
+for automated discovery of BPMN process models from event logs.
+
+Split Miner (Augusto et al., 2017/2019) produces simple process models
+with low branching complexity and consistently high and balanced fitness
+and precision, while being guaranteed to produce deadlock-free models
+with concurrency.
+
+This package implements Split Miner 1.0 with the following pipeline:
+
+1. DFG and loops discovery
+2. Concurrency discovery
+3. Edge filtering
+4. Split gateways discovery
+5. Join gateways discovery (via RPST / SPQR-tree)
+6. OR-joins minimization (via dominator tree)
+
+Features:
+
+- Pure Python --- no compiled extensions.
+- Implements Split Miner 1.0 from the original papers.
+- Uses SPQR-tree for correct RPST computation.
+- Typed package with PEP 561 support.
+- Requires Python 3.10 or later.
+
+
+Installation
+============
+
+You can install split-miner with ``pip``:
+
+::
+
+    pip install split-miner
+
+You may also install the latest source from the
+`split-miner GitHub repository`_.
+
+::
+
+    pip install git+https://github.com/imacat/split-miner.git
+
+
+Quick Start
+===========
+
+.. code-block:: python
+
+    from split_miner import BPMNModel, split_miner
+
+    # Create an event log (trace -> frequency)
+    traces: dict[tuple[str, ...], int] = {
+        ("a", "b", "c", "d"): 10,
+        ("a", "c", "b", "d"): 10,
+    }
+
+    # Discover a BPMN model
+    model: BPMNModel = split_miner(traces)
+
+    # Inspect the model
+    print(f"Tasks: {len(model.tasks)}")
+    print(f"Gateways: {len(model.gateways)}")
+    print(f"Edges: {len(model.edges)}")
+
+
+Parameters
+==========
+
+- **epsilon** (float, 0--1): Controls concurrency detection sensitivity.
+  Lower values require more balanced directly-follows frequencies to
+  detect concurrency.  Default: 0.33.
+- **eta** (float, 0--1): Controls edge filtering / retention.
+  Lower values retain more edges, resulting in higher fitness at the
+  cost of lower precision.  Default: 0.8.
+
+
+References
+==========
+
+- A. Augusto, R. Conforti, M. Dumas, M. La Rosa, and A. Polyvyanyy,
+  "Split Miner: Automated Discovery of Accurate and Simple Business
+  Process Models from Event Logs," *Knowledge and Information Systems*,
+  vol. 59, no. 2, pp. 251--284, 2019.
+  `doi:10.1007/s10115-018-1214-x`_
+
+- A. Augusto, R. Conforti, M. Dumas, M. La Rosa, and
+  A. Polyvyanyy, "Split Miner: Discovering Accurate and Simple
+  Business Process Models from Event Logs," *Proc. ICDM 2017*,
+  pp. 1--10, 2017. `doi:10.1109/ICDM.2017.9`_
+
+
+Acknowledgments
+===============
+
+This project was implemented from scratch in Python based on the
+original Split Miner papers.
+
+Development was assisted by `Claude Code`_ (Anthropic).
+
+
+Copyright
+=========
+
+ Copyright (c) 2026 imacat.
+
+ Licensed under the Apache License, Version 2.0 (the "License");
+ you may not use this file except in compliance with the License.
+ You may obtain a copy of the License at
+
+     http://www.apache.org/licenses/LICENSE-2.0
+
+ Unless required by applicable law or agreed to in writing, software
+ distributed under the License is distributed on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ See the License for the specific language governing permissions and
+ limitations under the License.
+
+
+Authors
+=======
+
+| imacat
+| imacat@mail.imacat.idv.tw
+| 2026/3/10
+
+.. _split-miner GitHub repository: https://github.com/imacat/split-miner
+.. _doi\:10.1007/s10115-018-1214-x: https://doi.org/10.1007/s10115-018-1214-x
+.. _doi\:10.1109/ICDM.2017.9: https://doi.org/10.1109/ICDM.2017.9
+.. _Claude Code: https://claude.com/claude-code