Automated product categorization

Chapters
1. Introduction
2. Service Architecture
3. Data
4. Training
5. Inference
6. Evaluation

#1 What is Skroutz.gr?
• Skroutz.gr is a marketplace & shopping assistant which
makes online shopping easier and more reliable
• It includes more than 11,000,000 products from 3,200
different e-shops
• On a monthly basis the website welcomes more than 8
million unique visitors ranking in the top positions in the
Greek Web

#1 Some Numbers
3,200
merchants
11
million products
270 mil.
pageviews /
mo
1.1 mil.
searches/day
33 mil.
sessions/mo

#1 The Problem
• Each day we collect thousands of new products by
downloading e-shop feeds (XML, CSV etc. - product
catalogs)
• We want to categorize incoming product payloads as
provided by eshops to the most relevant categories in
Skroutz category tree taxonomy with the minimum human
intervention.
- Difficult
- Important

#1 Why Difficult?
• Many leaf categories in
Skroutz taxonomy (>2k)
• Sibling categories
(subjective categorization)
• Misleading product titles
and shop-categories from
shops

#1 Why Important?
Robot MO collects
products from shop
feeds and stores them
to DB
Megatron category
classifier categorizes
products to the correct
category
Tron groups similar
products to entities
called SKUs to be
ready for indexing
Elasticsearch indexes
products to be
searchable from user
interface

#1 Facts
•Merchants send more than ~15k new products every day in
Skroutz!!!
•2.3k unique leaf categories in our category tree (taxonomy)
•Manual “move-to-category” action:
- Costs ~7.8s on average for content managers
- Subjective decisions may add extra overhead

#1 Old Solution - Overview
•Use Elasticseach to match specific product attributes:
- PN (manufacturer part number)
- Name
- Shop-category
•Aggregate matches and group by categories
•Normalize results and use custom weights to calculate a score
•Take Top-K results

#1 Old Solution - Limitations
•Plain cosine similarity distance on TF/IDF weights:
- No learning feedback loop
- No advanced statistics utilization (e.g. correlation between price
value and text features)
•No easy way to tune custom weights applied on final scoring
•Heuristics don’t take into account category specific context
•Heuristics don’t take into account word level context. E.g.
word “samsung” is followed by word “galaxy” most of the time
and then probably follows a model number.

#1 Old Solution - Good Parts
•Simple solution (except for custom scoring stuff)
•Easy to debug
•Easy to deploy
•Online

#1 New Solution - “Megatron”

#1 Overview
•Approach problem as a supervised learning task
•Rely on probabilities to obtain a meaningful score
•Use more features from multiple sources and use datasets
•Learn new patterns and relations by training
•Measure performance on dataset splits
•Use a microservice to serve classification requests
•Apply threshold for low confidence results

#2 Service Architecture
1.Training Phase
2.Inference Phase
3.APIs

#2.1 Training Phase
1. Export dataset (product features labeled with category_id) and upload to Swift
2. Download specific dataset version in “training VM”
3. Start a training session using a train/val split from dataset
4. Save best performing model params snapshot (based on validation set loss)
5. Compress and upload model params to Swift container

#2.2 Inference Phase
1. Application Part: Send classification request
upon new product arrivals:
- Kafka producer (asynchronous request)
- Megatron Client HTTP synchronous
requests (2nd alternative)
2. Category Classifier Microservice Part:
- Pop messages from stream (Kafka
consumer)
- Dispatch messages to in-memory Neural
Network instance
- Fetch predictions (scores) and post-back
to Core Application API endpoint

#2.3 APIs
1. Megatron microservice internal API
- Common API (wraps Keras API)
- Basic methods:
✓ build
✓ train
✓ save
✓ load
✓ predict
- CLI commands

#2.3 APIs(2)
1. Skroutz Application Ecosystem (Ruby client)
- Megatron::Client
✓ Issues requests to microservice
- Megatron DB model
✓ Stores prediction results
- ApiController endpoint
✓ Receives callbacks from microservice

#3 Data
•Product attribute values (potential features)
Product
Name
Shop manufacturer
Part number
EAN
Price
Shop category
...
Samsung TV 32'' DF324 (PNDFD22) Full HD Black NEW
Αρχική > Ηλεκτρονικά > Τηλεοράσεις
PNDFD22
300 €

#3 Data(2)
•Training Dataset - Raw Features
Image
Numerical
Categorical
Label
Text

#3 Data(3)
•Preprocessing
- Text
- Numerical
- Categorical
- Labels
X
y

#3 Preprocessing - Text
• Our best solution involves “Word Vectors”
• Steps to prepare for word vectors:
- Learn a words Vocabulary (mapping of words to numeric id)
- Transform text sentences to Sequences of ids based on Vocabulary
- Decide a representative sequence length (E.g. 60 words)
- Apply zero padding (pre or post) and truncation to maintain a fixed length

#3 Preprocessing - Text(2)
• Use of Pretrained Embeddings (see W2Vec, FastText, GloVe etc.)
• We use FastText library with skipgram algorithm (unsupervised)
- https://fasttext.cc/docs/en/unsupervised-tutorial.html

#3 Preprocessing - Text(3)
• Embeddings:
- Outputs 100 dim Vector
- Total 1,500,000 rows (vocab)
• 2 versions (Name, Shop-category)

#3 Preprocessing - Numerical
• “Pricevat” and “Name Length” values
• Apply Standard Scaling

#3 Preprocessing - Categorical
• All discrete value attributes/features:
- shop_id
- matching Product PNs category_id list
• One-Hot encoding:

#3 Label Encoding
• “category_id “ values are the “true” labels which should be learned by NN
• One-Hot encoding
• OR just use IDs and rely to “Keras” conventions (E.g. use an internal sparse categorical
representation to save huge amounts of RAM)

#4 Training
1.Basic Concepts
2.Model Architecture
3.Training “In Action”

#4.1 Basic Concepts
•Objective:
- Find a combination of mathematical functions and a set of
corresponding params to maximize prediction accuracy (or minimize
error rate).
- Ensure that the above generalizes well for production.
- Learn params in an acceptable time window.
•Experiment with Neural Network architectures
•GPUS to the rescue (speedup x10)

#4.1 Basic Concepts(2)
•Loss function
- Categorical Crossentropy
•Optimizer
- Adam (Gradient Descent)
•Hyper-params
- Mini-Batch Size
- Learning Rate
- Epochs

#4.1 Validation
•Why?
- Simulate unseen data
- Compare different:
✓ training methods
✓ hyper -params
- Avoid Overfitting
•Should be representative
•Validation Strategy
- 10% of whole Dataset
- Stratification on Categories

#4.2 Model Architecture
•Hybrid End-to-End architecture
•4 branches (4 input vectors):
A. Name Features Branch
B. Shop-Category Features Branch
C. Basic Features Branch (Numerics, Categorical)
D. Matching PNs Branch (Categorical)
Text

#4.2 Text Branches
• Inspired by “Embed, Encode, Attend, Predict”
- https://explosion.ai/blog/deep-learning-formula-nlp
• Each of “name” and “shop-category” sequence flows through:
- 1 x Embeddings Layer
- 1 x Bi-LSTM Encoder
- 1 x Attention Module
- 1 x LSTM Encoder

#4.2 Text Branches - Why LSTM?
• LSTM stands for “Long Short Term Memory” Layer (Encoder):
- Memory Cells / Captures context
- Propagates signal from previous words to the next in a Sequence
- 2 Stacked Layers performed better in our experiments
- 128 dimension output vector
- https://colah.github.io/posts/2015-08-Understanding-LSTMs/
128dim
#cells = sequence length

#4.2 Text Branches - Why pay Attention?
• Attention Mechanism:
- Controll how much signal should be propagated to next layers
- https://distill.pub/2016/augmented-rnns/

#4.2 Other Branches
• Basic Features Branch
- Inputs a concatenation of basic feats
- 1 Dense layer with #classes output
- ReLU activation
• Matching PNs Branch
- Inputs a concatenation of PN feats
- Short-circuited to final layer
InputVector
#classes
(~2kforSkroutz)

#4.2 Final Layer
• Merging Layer
- Concatenates all 4 branches outputs
- softmax activation
- Output: probabilities for each class

#4.2 Model Architecture
• Model Capacity/Complexity:

#4.3 Training In Action - Model Selection
•Conducted 100s of experiments with different combinations
of features, layers, modules (e.g. Embeddings, Bag of Words,
TF/IDF, LSTM, etc.)
•10s of Ablations studies: remove specific features to see how
performance is affected
•Read many papers and applied some common tricks (Bi-LSTM,
AdaptivePooling etc.)
•It is an alchemy!

#4.3 Training In Action - Tools
•Training Scheduler Process runs weekly
•CLI training commands
- CUDA_VISIBLE_DEVICES=1 python -m category_classifier.cli scrooge --model end2end --train --epochs
8 --batch_size 128
•Model Versioning
- E.g. “skroutz_models_2018_09_01_v1.tar.gz”

#4.3 Training In Action
Training run output example:
GPU monitoring:

#4.3 Training In Action
Learning Curves (Tensorboard):
Current best
Previous Arch Current bestPrevious Arch

#5 Inference
1.Inference Pipeline
2.Inference API
3.Production

#5.1 Inference Pipeline
•Online execution:
- preprocessing
- vectorization
- Prediction
•Utilized by CategoryClassifier Class
- Wrapper of external API
•Utilize scikit-learn Pipelines
- http://scikit-learn.org/stable/modules/generated/sklearn.pipeline.Pipeline.html

#5.2 Inference API
• REPL
• Kafka Worker
• Flask App

#5.3 Production
•x2 inference VMs
- inference1.skroutz.gr, inference2.skroutz.gr (Kafka Workers)
•x2 Flavors (Greece, UK)
•Grafana Monitoring for Kafka Part

#6 Evaluation
•More than 6% error rate reduction overall in Skroutz!
•Currently, more than ~2 content-editor hours saved per day in
Skroutz (this is scaling)!
•Move operations from list with “uncategorized” products
reduced significantly (by an order of magnitude)!

#6 Performance Summary
Success Rate Failure Rate No Prediction Rate
Megatron Old Megatron Old Megatron Old
Skroutz (GR)
2.3k categories
90.10% 82.6% 7.9% 13.8% 2% 3.5%
91.85% 85.7% 8.14% 14.32% N/A N/A
Scrooge (UK)
350 categories
87.56% 38.9% 2.5% 26.24% 9.9% 58.48%
97.1% 93.67% 2.8% 6.32% N/A N/A

#Future Improvements
• Utilize Image Features (in End-To-End model)
• Utilize Entity Recognition to extract more features
• Find ways to utilize more features (color, sizes etc.)
• Categorical Self-Trained Embeddings
• Experiment with newer solutions like “Transformer”

#Contact Info
Andreas Loupasakis
• Email: alup@skroutz.gr
• Kaggle: https://www.kaggle.com/andreaslup
• Twitter: https://twitter.com/andy_lupo
• LinkedIn: https://www.linkedin.com/in/andreas-loupasakis-06399a47

Automated product categorization

More Related Content

Similar to Automated product categorization

More from Warply

Recently uploaded

Automated product categorization