├── .gitignore
├── LICENSE
├── README.md
├── data
└── ecomm
│ └── OnlineRetail_sessions.pkl
├── notebooks
├── Analyze_HPO_results.ipynb
└── Explore_Online_Retail_Dataset.ipynb
├── recsys
├── __init__.py
├── data.py
├── metrics.py
├── models.py
└── utils.py
├── requirements.txt
├── requirements3.6.txt
├── scripts
├── baseline_analysis.py
├── setup_ray_cluster.py
├── train_w2v_with_logging.py
└── tune_w2v_with_ray.py
└── setup.py
/.gitignore:
--------------------------------------------------------------------------------
1 | # Byte-compiled / optimized / DLL files
2 | __pycache__/
3 | *.py[cod]
4 | *$py.class
5 |
6 | # C extensions
7 | *.so
8 |
9 | # Distribution / packaging
10 | .Python
11 | build/
12 | develop-eggs/
13 | dist/
14 | downloads/
15 | eggs/
16 | .eggs/
17 | lib/
18 | lib64/
19 | parts/
20 | sdist/
21 | var/
22 | wheels/
23 | share/python-wheels/
24 | *.egg-info/
25 | .installed.cfg
26 | *.egg
27 | MANIFEST
28 |
29 | # PyInstaller
30 | # Usually these files are written by a python script from a template
31 | # before PyInstaller builds the exe, so as to inject date/other infos into it.
32 | *.manifest
33 | *.spec
34 |
35 | # Installer logs
36 | pip-log.txt
37 | pip-delete-this-directory.txt
38 |
39 | # Unit test / coverage reports
40 | htmlcov/
41 | .tox/
42 | .nox/
43 | .coverage
44 | .coverage.*
45 | .cache
46 | nosetests.xml
47 | coverage.xml
48 | *.cover
49 | *.py,cover
50 | .hypothesis/
51 | .pytest_cache/
52 | cover/
53 |
54 | # Translations
55 | *.mo
56 | *.pot
57 |
58 | # Django stuff:
59 | *.log
60 | local_settings.py
61 | db.sqlite3
62 | db.sqlite3-journal
63 |
64 | # Flask stuff:
65 | instance/
66 | .webassets-cache
67 |
68 | # Scrapy stuff:
69 | .scrapy
70 |
71 | # Sphinx documentation
72 | docs/_build/
73 |
74 | # PyBuilder
75 | .pybuilder/
76 | target/
77 |
78 | # Jupyter Notebook
79 | .ipynb_checkpoints
80 |
81 | # IPython
82 | profile_default/
83 | ipython_config.py
84 |
85 | # pyenv
86 | # For a library or package, you might want to ignore these files since the code is
87 | # intended to run in multiple environments; otherwise, check them in:
88 | # .python-version
89 |
90 | # pipenv
91 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
92 | # However, in case of collaboration, if having platform-specific dependencies or dependencies
93 | # having no cross-platform support, pipenv may install dependencies that don't work, or not
94 | # install all needed dependencies.
95 | #Pipfile.lock
96 |
97 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow
98 | __pypackages__/
99 |
100 | # Celery stuff
101 | celerybeat-schedule
102 | celerybeat.pid
103 |
104 | # SageMath parsed files
105 | *.sage.py
106 |
107 | # Environments
108 | .env
109 | .venv
110 | env/
111 | venv/
112 | ENV/
113 | env.bak/
114 | venv.bak/
115 |
116 | # Spyder project settings
117 | .spyderproject
118 | .spyproject
119 |
120 | # Rope project settings
121 | .ropeproject
122 |
123 | # mkdocs documentation
124 | /site
125 |
126 | # mypy
127 | .mypy_cache/
128 | .dmypy.json
129 | dmypy.json
130 |
131 | # Pyre type checker
132 | .pyre/
133 |
134 | # pytype static type analyzer
135 | .pytype/
136 |
137 | # Cython debug symbols
138 | cython_debug/
139 |
140 | # data directory
141 | #data/
142 |
--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
1 | Apache License
2 | Version 2.0, January 2004
3 | http://www.apache.org/licenses/
4 |
5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
6 |
7 | 1. Definitions.
8 |
9 | "License" shall mean the terms and conditions for use, reproduction,
10 | and distribution as defined by Sections 1 through 9 of this document.
11 |
12 | "Licensor" shall mean the copyright owner or entity authorized by
13 | the copyright owner that is granting the License.
14 |
15 | "Legal Entity" shall mean the union of the acting entity and all
16 | other entities that control, are controlled by, or are under common
17 | control with that entity. For the purposes of this definition,
18 | "control" means (i) the power, direct or indirect, to cause the
19 | direction or management of such entity, whether by contract or
20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the
21 | outstanding shares, or (iii) beneficial ownership of such entity.
22 |
23 | "You" (or "Your") shall mean an individual or Legal Entity
24 | exercising permissions granted by this License.
25 |
26 | "Source" form shall mean the preferred form for making modifications,
27 | including but not limited to software source code, documentation
28 | source, and configuration files.
29 |
30 | "Object" form shall mean any form resulting from mechanical
31 | transformation or translation of a Source form, including but
32 | not limited to compiled object code, generated documentation,
33 | and conversions to other media types.
34 |
35 | "Work" shall mean the work of authorship, whether in Source or
36 | Object form, made available under the License, as indicated by a
37 | copyright notice that is included in or attached to the work
38 | (an example is provided in the Appendix below).
39 |
40 | "Derivative Works" shall mean any work, whether in Source or Object
41 | form, that is based on (or derived from) the Work and for which the
42 | editorial revisions, annotations, elaborations, or other modifications
43 | represent, as a whole, an original work of authorship. For the purposes
44 | of this License, Derivative Works shall not include works that remain
45 | separable from, or merely link (or bind by name) to the interfaces of,
46 | the Work and Derivative Works thereof.
47 |
48 | "Contribution" shall mean any work of authorship, including
49 | the original version of the Work and any modifications or additions
50 | to that Work or Derivative Works thereof, that is intentionally
51 | submitted to Licensor for inclusion in the Work by the copyright owner
52 | or by an individual or Legal Entity authorized to submit on behalf of
53 | the copyright owner. For the purposes of this definition, "submitted"
54 | means any form of electronic, verbal, or written communication sent
55 | to the Licensor or its representatives, including but not limited to
56 | communication on electronic mailing lists, source code control systems,
57 | and issue tracking systems that are managed by, or on behalf of, the
58 | Licensor for the purpose of discussing and improving the Work, but
59 | excluding communication that is conspicuously marked or otherwise
60 | designated in writing by the copyright owner as "Not a Contribution."
61 |
62 | "Contributor" shall mean Licensor and any individual or Legal Entity
63 | on behalf of whom a Contribution has been received by Licensor and
64 | subsequently incorporated within the Work.
65 |
66 | 2. Grant of Copyright License. Subject to the terms and conditions of
67 | this License, each Contributor hereby grants to You a perpetual,
68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable
69 | copyright license to reproduce, prepare Derivative Works of,
70 | publicly display, publicly perform, sublicense, and distribute the
71 | Work and such Derivative Works in Source or Object form.
72 |
73 | 3. Grant of Patent License. Subject to the terms and conditions of
74 | this License, each Contributor hereby grants to You a perpetual,
75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable
76 | (except as stated in this section) patent license to make, have made,
77 | use, offer to sell, sell, import, and otherwise transfer the Work,
78 | where such license applies only to those patent claims licensable
79 | by such Contributor that are necessarily infringed by their
80 | Contribution(s) alone or by combination of their Contribution(s)
81 | with the Work to which such Contribution(s) was submitted. If You
82 | institute patent litigation against any entity (including a
83 | cross-claim or counterclaim in a lawsuit) alleging that the Work
84 | or a Contribution incorporated within the Work constitutes direct
85 | or contributory patent infringement, then any patent licenses
86 | granted to You under this License for that Work shall terminate
87 | as of the date such litigation is filed.
88 |
89 | 4. Redistribution. You may reproduce and distribute copies of the
90 | Work or Derivative Works thereof in any medium, with or without
91 | modifications, and in Source or Object form, provided that You
92 | meet the following conditions:
93 |
94 | (a) You must give any other recipients of the Work or
95 | Derivative Works a copy of this License; and
96 |
97 | (b) You must cause any modified files to carry prominent notices
98 | stating that You changed the files; and
99 |
100 | (c) You must retain, in the Source form of any Derivative Works
101 | that You distribute, all copyright, patent, trademark, and
102 | attribution notices from the Source form of the Work,
103 | excluding those notices that do not pertain to any part of
104 | the Derivative Works; and
105 |
106 | (d) If the Work includes a "NOTICE" text file as part of its
107 | distribution, then any Derivative Works that You distribute must
108 | include a readable copy of the attribution notices contained
109 | within such NOTICE file, excluding those notices that do not
110 | pertain to any part of the Derivative Works, in at least one
111 | of the following places: within a NOTICE text file distributed
112 | as part of the Derivative Works; within the Source form or
113 | documentation, if provided along with the Derivative Works; or,
114 | within a display generated by the Derivative Works, if and
115 | wherever such third-party notices normally appear. The contents
116 | of the NOTICE file are for informational purposes only and
117 | do not modify the License. You may add Your own attribution
118 | notices within Derivative Works that You distribute, alongside
119 | or as an addendum to the NOTICE text from the Work, provided
120 | that such additional attribution notices cannot be construed
121 | as modifying the License.
122 |
123 | You may add Your own copyright statement to Your modifications and
124 | may provide additional or different license terms and conditions
125 | for use, reproduction, or distribution of Your modifications, or
126 | for any such Derivative Works as a whole, provided Your use,
127 | reproduction, and distribution of the Work otherwise complies with
128 | the conditions stated in this License.
129 |
130 | 5. Submission of Contributions. Unless You explicitly state otherwise,
131 | any Contribution intentionally submitted for inclusion in the Work
132 | by You to the Licensor shall be under the terms and conditions of
133 | this License, without any additional terms or conditions.
134 | Notwithstanding the above, nothing herein shall supersede or modify
135 | the terms of any separate license agreement you may have executed
136 | with Licensor regarding such Contributions.
137 |
138 | 6. Trademarks. This License does not grant permission to use the trade
139 | names, trademarks, service marks, or product names of the Licensor,
140 | except as required for reasonable and customary use in describing the
141 | origin of the Work and reproducing the content of the NOTICE file.
142 |
143 | 7. Disclaimer of Warranty. Unless required by applicable law or
144 | agreed to in writing, Licensor provides the Work (and each
145 | Contributor provides its Contributions) on an "AS IS" BASIS,
146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
147 | implied, including, without limitation, any warranties or conditions
148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
149 | PARTICULAR PURPOSE. You are solely responsible for determining the
150 | appropriateness of using or redistributing the Work and assume any
151 | risks associated with Your exercise of permissions under this License.
152 |
153 | 8. Limitation of Liability. In no event and under no legal theory,
154 | whether in tort (including negligence), contract, or otherwise,
155 | unless required by applicable law (such as deliberate and grossly
156 | negligent acts) or agreed to in writing, shall any Contributor be
157 | liable to You for damages, including any direct, indirect, special,
158 | incidental, or consequential damages of any character arising as a
159 | result of this License or out of the use or inability to use the
160 | Work (including but not limited to damages for loss of goodwill,
161 | work stoppage, computer failure or malfunction, or any and all
162 | other commercial damages or losses), even if such Contributor
163 | has been advised of the possibility of such damages.
164 |
165 | 9. Accepting Warranty or Additional Liability. While redistributing
166 | the Work or Derivative Works thereof, You may choose to offer,
167 | and charge a fee for, acceptance of support, warranty, indemnity,
168 | or other liability obligations and/or rights consistent with this
169 | License. However, in accepting such obligations, You may act only
170 | on Your own behalf and on Your sole responsibility, not on behalf
171 | of any other Contributor, and only if You agree to indemnify,
172 | defend, and hold each Contributor harmless for any liability
173 | incurred by, or claims asserted against, such Contributor by reason
174 | of your accepting any such warranty or additional liability.
175 |
176 | END OF TERMS AND CONDITIONS
177 |
178 | APPENDIX: How to apply the Apache License to your work.
179 |
180 | To apply the Apache License to your work, attach the following
181 | boilerplate notice, with the fields enclosed by brackets "[]"
182 | replaced with your own identifying information. (Don't include
183 | the brackets!) The text should be enclosed in the appropriate
184 | comment syntax for the file format. We also recommend that a
185 | file or class name and description of purpose be included on the
186 | same "printed page" as the copyright notice for easier
187 | identification within third-party archives.
188 |
189 | Copyright [yyyy] [name of copyright owner]
190 |
191 | Licensed under the Apache License, Version 2.0 (the "License");
192 | you may not use this file except in compliance with the License.
193 | You may obtain a copy of the License at
194 |
195 | http://www.apache.org/licenses/LICENSE-2.0
196 |
197 | Unless required by applicable law or agreed to in writing, software
198 | distributed under the License is distributed on an "AS IS" BASIS,
199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
200 | See the License for the specific language governing permissions and
201 | limitations under the License.
202 |
--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # Session-based Recommender Systems
2 |
3 | This repo accompanies the Cloudera Fast Forward report [Session-based Recommender Systems](https://session-based-recommenders.fastforwardlabs.com/). It provides small library to train Word2Vec as a means of learning product or item representations in the context of user sessions (browsing histories, transaction histories, music playlists, etc.). These dense representations can then be used for item recommendation. We formulate this under the Next Event Prediction task, that is, given a user's recent interaction, predict the next item they interact with (click on, purchase, listen to, etc.).
4 |
5 | Instructions are given both for general use (on a laptop, say), and for Cloudera CML and CDSW. We'll first describe what's here, then go through how to run everything.
6 |
7 | ## Structure
8 | ```
9 | .
10 | ├── data # This folder contains starter data.
11 | ├── scripts # This contains scripts for *doing* things -- training models, analysing results.
12 | ├── notebooks # This contains Jupyter notebooks that accompany the report and demonstrate basic usage.
13 | └── recsys # A small library of useful functions.
14 | ```
15 | Let's examine each of the important folders in turn.
16 |
17 |
18 | ### `recsys`
19 | ```
20 | ├── data.py # Contains functions for loading and processing data into sessions
21 | ├── metrics.py # Contains metrics for evaluation
22 | ├── models.py # Contains wrappers for training Word2Vec both alone and with Ray Tune
23 | └── utils.py # Helper functions for serialization and I/O
24 | ```
25 |
26 |
27 | ### `scripts`
28 | ```
29 | ├── baseline_analysis.py
30 | ├── setup_ray_cluster.py
31 | ├── train_w2v_with_logging.py
32 | └── tune_w2v_with_ray.py
33 | ```
34 | An overview of what each of these scripts does is discussed below.
35 |
36 | ### `notebooks`
37 | ```
38 | ├── Analyze_HPO_results.ipynb
39 | └── Explore_Online_Retail_Dataset.ipynb
40 | ```
41 | These notebooks provide additional exploration and analysis. Please note that `Analyze_HPO_results.ipynb` is expressly for demonstration purposes as HPO output results explored within are not included in this repo.
42 |
43 | ## Learning representations for session-based recommendations
44 | To go from a fresh clone of the repo to the final state, follow these instructions in order.
45 |
46 | ### Installation
47 | The code and applications within were developed against Python 3.8.8, and are likely also to function with more recent versions of Python.
48 |
49 | To install dependencies, first create and activate a new virtual environment through your preferred means, then pip install from the requirements file. I recommend:
50 |
51 | ```
52 | python3 -m venv .venv
53 | source .venv/bin/activate
54 | pip install -r requirements.txt
55 | ```
56 |
57 | In CML or CDSW, no virtual env is necessary. Instead, inside a Python 3 session (with at least 2 vCPU / 4 GiB Memory), simply run
58 |
59 | ```
60 | !pip3 install -r requirements.txt # notice `pip3`, not `pip`
61 | ```
62 |
63 | Note: if your session has an older Python image (3.6) use the alternative `requirements3.6.txt`:
64 | ```
65 | !pip3 install -r requirements3.6.txt
66 | ```
67 |
68 | ### Data
69 |
70 | While we explored several datasets (and code exists in `recsys/data.py` to interact with those datasets), the analysis in this repo is focused on the [Online Retail](https://www.kaggle.com/vijayuv/onlineretail) dataset. This dataset is open source though you will need to create an account on Kaggle before downloading the data. In this repo we include a version of this dataset post-processed into customer sessions. These sessions represent all customer transactions from a UK-based online boutique selling specialty gifts collected between 12/01/2010 and 12/09/2011. In total there are purchase histories for 4,372 customers and 3,684 unique products.
71 |
72 | ### Model training and analysis
73 |
74 | The `scripts` directory contains scripts to train models in various formats and analyze results. Here we provide a high-level overview:
75 |
76 | * `scripts/baseline_analysis.py`: a common baseline for recommendation systems is to simply recommend the most popular items. This script computes the "Association Rules" baseline which considers how frequently each item co-occurrs with all other items in a session for each session in the training set.
77 | * `scripts/train_w2v_with_logging.py`: This script trains Gensim's implementation of the Word2Vec algorithm to learn representations for each item in a session. Identifying "similar" items then serves as the method for generating recommendations. Includes callbacks for monitoring metrics (Recall@K, training loss) as a function of training time (epochs).
78 | * `scripts/tune_w2v_with_ray.py`: The Word2Vec algorithm has a large hyperparameter space and the default values are subpar for the task of generating good item representations for recommendation systems. This scripts performs hyperparameter optimization (HPO) with [Ray Tune](https://docs.ray.io/en/master/tune/index.html).
79 | * [CDSW/CML only] `setup_ray_cluster.py`: Hyperparameter optimization can be computationally expensive but this expense can be mitigated, in part, through distribution. This script initializes (and tears down) a Ray Cluster for distributed hyperparameter optimization. If using, follow the instructions in this script to setup the cluster, then run `tune_w2v_with_ray.py` with the appropriate arguments, and finally shutdown the cluster after HPO is complete.
80 |
81 |
82 | These scripts are not intended to be run in any particular order (with the exception noted above). Instead, they provide functionality for different use cases. To run scripts, follow this procedure in the terminal or a Session with at least 1vCPUs and 2GiBs of memory:
83 |
84 | ```
85 | !python3 scripts/baseline_analysis.py
86 | !python3 scripts/train_w2v_with_logging.py # see optional arguments
87 | !python3 scriptstune_w2v_with_ray.py # see optional arguments for distributed HPO
88 | ```
89 |
90 |
--------------------------------------------------------------------------------
/data/ecomm/OnlineRetail_sessions.pkl:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/fastforwardlabs/session_based_recommenders/c438dd1334fcefc6bedea69b0cd67f779a5de5d3/data/ecomm/OnlineRetail_sessions.pkl
--------------------------------------------------------------------------------
/notebooks/Explore_Online_Retail_Dataset.ipynb:
--------------------------------------------------------------------------------
1 | {
2 | "cells": [
3 | {
4 | "cell_type": "markdown",
5 | "metadata": {},
6 | "source": [
7 | "# The \"Online Retail\" Dataset \n",
8 | "\n",
9 | "This is a transnational data set which contains all purchase transactions occurring between 01/12/2010 and 09/12/2011 for a UK-based and registered non-store online retail. The company mainly sells unique all-occasion gifts. Many customers of the company are wholesalers. \n",
10 | "\n",
11 | "The dataset is composed of the following columns:\n",
12 | "\n",
13 | "\n",
14 | "* **InvoiceNo**: Invoice number. Nominal, a 6-digit integral number uniquely assigned to each transaction. If this code starts with letter 'c', it indicates a cancellation.\n",
15 | "* **StockCode**: Product (item) code. Nominal, a 5-digit integral number uniquely assigned to each distinct product.\n",
16 | "* **Description**: Product (item) name. Nominal.\n",
17 | "* **Quantity**: The quantities of each product (item) per transaction. Numeric.\n",
18 | "* **InvoiceDate**: Invice Date and time. Numeric, the day and time when each transaction was generated.\n",
19 | "* **UnitPrice**: Unit price. Numeric, Product price per unit in sterling.\n",
20 | "* **CustomerID**: Customer number. Nominal, a 5-digit integral number uniquely assigned to each customer.\n",
21 | "* **Country**: Country name. Nominal, the name of the country where each customer resides."
22 | ]
23 | },
24 | {
25 | "cell_type": "code",
26 | "execution_count": 22,
27 | "metadata": {},
28 | "outputs": [],
29 | "source": [
30 | "import pandas as pd\n",
31 | "import numpy as np\n",
32 | "import matplotlib.pyplot as plt\n",
33 | "import seaborn as sns\n",
34 | "\n",
35 | "from recsys.data import load_original_ecomm, preprocess_ecomm, construct_session_sequences "
36 | ]
37 | },
38 | {
39 | "cell_type": "code",
40 | "execution_count": 23,
41 | "metadata": {},
42 | "outputs": [],
43 | "source": [
44 | "import seaborn as sns\n",
45 | "plt.style.use(\"seaborn-white\")\n",
46 | "cldr_colors = ['#00b6b5', '#f7955b','#6c8cc7', '#828282']#\n",
47 | "cldr_green = '#a4d65d'\n",
48 | "color_palette = \"viridis\""
49 | ]
50 | },
51 | {
52 | "cell_type": "markdown",
53 | "metadata": {},
54 | "source": [
55 | "## Getting the dataset\n",
56 | "\n",
57 | "We obtained the Online Retail dataset from the Kaggle website found [here](https://www.kaggle.com/vijayuv/onlineretail). \n",
58 | "The data is open source but you will need to register on Kaggle's website before downloading. \n",
59 | "\n",
60 | "Once downloaded, we created some helper functions for opening and processing this dataset. "
61 | ]
62 | },
63 | {
64 | "cell_type": "code",
65 | "execution_count": 39,
66 | "metadata": {},
67 | "outputs": [],
68 | "source": [
69 | "df = load_original_ecomm()"
70 | ]
71 | },
72 | {
73 | "cell_type": "code",
74 | "execution_count": 25,
75 | "metadata": {},
76 | "outputs": [
77 | {
78 | "data": {
79 | "text/html": [
80 | "
"
549 | ]
550 | },
551 | "metadata": {},
552 | "output_type": "display_data"
553 | }
554 | ],
555 | "source": [
556 | "plt.style.use(\"seaborn-white\")\n",
557 | "\n",
558 | "# Number of unique customer IDs\n",
559 | "product_counts = df.groupby(['StockCode']).count()['InvoiceNo'].values\n",
560 | "\n",
561 | "fig = plt.figure(figsize=(8,6))\n",
562 | "plt.yticks(fontsize=14)\n",
563 | "plt.xticks(fontsize=14)\n",
564 | "\n",
565 | "plt.semilogy(sorted(product_counts))\n",
566 | "plt.ylabel(\"Product counts\", fontsize=16);\n",
567 | "plt.xlabel(\"Product index\", fontsize=16);\n",
568 | "\n",
569 | "plt.tight_layout()"
570 | ]
571 | },
572 | {
573 | "cell_type": "markdown",
574 | "metadata": {},
575 | "source": [
576 | "The left side of the figure corresponds to products that are not very popular (because they aren't purchased very often), while the far right side indicates that some products are *extremely* popular and have been purchased hundreds of times. \n",
577 | "\n",
578 | "### Customer session lengths \n",
579 | "\n",
580 | "We define a customer's \"session\" as all the products they purchased in each transaction, in the order in which they were purchased (ordered InvoiceDate). We can then examine statistics regarding the length of these sessions. Below is a boxplot of all customer session lengths. "
581 | ]
582 | },
583 | {
584 | "cell_type": "code",
585 | "execution_count": 78,
586 | "metadata": {},
587 | "outputs": [
588 | {
589 | "data": {
590 | "image/png": "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\n",
591 | "text/plain": [
592 | "