Classifying Duplicate Questions from Quora with Keras

Introduction

On this submit we are going to use Keras to categorise duplicated questions from Quora.
The dataset first appeared within the Kaggle competitors Quora Query Pairs and consists of roughly 400,000 pairs of questions together with a column indicating if the query pair is taken into account a reproduction.

Our implementation is impressed by the Siamese Recurrent Structure, with modifications to the similarity
measure and the embedding layers (the unique paper makes use of pre-trained phrase vectors). Utilizing this type
of structure dates again to 2005 with Le Cun et al and is beneficial for
verification duties. The thought is to study a perform that maps enter patterns right into a
goal house such {that a} similarity measure within the goal house approximates
the “semantic” distance within the enter house.

After the competitors, Quora additionally described their method to this downside on this weblog submit.

Dowloading knowledge

Information might be downloaded from the Kaggle dataset webpage
or from Quora’s launch of the dataset:

library(keras)
quora_data <- get_file(
  "quora_duplicate_questions.tsv",
  "https://qim.ec.quoracdn.web/quora_duplicate_questions.tsv"
)

We’re utilizing the Keras get_file() perform in order that the file obtain is cached.

Studying and preprocessing

We are going to first load knowledge into R and do some preprocessing to make it simpler to
embody within the mannequin. After downloading the information, you may learn it
utilizing the readr read_tsv() perform.

We are going to create a Keras tokenizer to remodel every phrase into an integer
token. We can even specify a hyperparameter of our mannequin: the vocabulary dimension.
For now let’s use the 50,000 commonest phrases (we’ll tune this parameter later).
The tokenizer can be match utilizing all distinctive questions from the dataset.

tokenizer <- text_tokenizer(num_words = 50000)
tokenizer %>% fit_text_tokenizer(distinctive(c(df$question1, df$question2)))

Let’s save the tokenizer to disk to be able to use it for inference later.

save_text_tokenizer(tokenizer, "tokenizer-question-pairs")

We are going to now use the textual content tokenizer to remodel every query into an inventory
of integers.

question1 <- texts_to_sequences(tokenizer, df$question1)
question2 <- texts_to_sequences(tokenizer, df$question2)

Let’s check out the variety of phrases in every query. It will helps us to
resolve the padding size, one other hyperparameter of our mannequin. Padding the sequences normalizes them to the identical dimension in order that we will feed them to the Keras mannequin.

80% 90% 95% 99% 
 14  18  23  31 

We will see that 99% of questions have at most size 31 so we’ll select a padding
size between 15 and 30. Let’s begin with 20 (we’ll additionally tune this parameter later).
The default padding worth is 0, however we’re already utilizing this worth for phrases that
don’t seem throughout the 50,000 most frequent, so we’ll use 50,001 as an alternative.

question1_padded <- pad_sequences(question1, maxlen = 20, worth = 50000 + 1)
question2_padded <- pad_sequences(question2, maxlen = 20, worth = 50000 + 1)

Now we have now completed the preprocessing steps. We are going to now run a easy benchmark
mannequin earlier than shifting on to the Keras mannequin.

Easy benchmark

Earlier than creating an advanced mannequin let’s take a easy method.
Let’s create two predictors: share of phrases from question1 that
seem within the question2 and vice-versa. Then we are going to use a logistic
regression to foretell if the questions are duplicate.

perc_words_question1 <- map2_dbl(question1, question2, ~imply(.x %in% .y))
perc_words_question2 <- map2_dbl(question2, question1, ~imply(.x %in% .y))

df_model <- knowledge.body(
  perc_words_question1 = perc_words_question1,
  perc_words_question2 = perc_words_question2,
  is_duplicate = df$is_duplicate
) %>%
  na.omit()

Now that now we have our predictors, let’s create the logistic mannequin.
We are going to take a small pattern for validation.

val_sample <- pattern.int(nrow(df_model), 0.1*nrow(df_model))
logistic_regression <- glm(
  is_duplicate ~ perc_words_question1 + perc_words_question2, 
  household = "binomial",
  knowledge = df_model[-val_sample,]
)
abstract(logistic_regression)

Name:
glm(method = is_duplicate ~ perc_words_question1 + perc_words_question2, 
    household = "binomial", knowledge = df_model[-val_sample, ])

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.5938  -0.9097  -0.6106   1.1452   2.0292  

Coefficients:
                      Estimate Std. Error z worth Pr(>|z|)    
(Intercept)          -2.259007   0.009668 -233.66   <2e-16 ***
perc_words_question1  1.517990   0.023038   65.89   <2e-16 ***
perc_words_question2  1.681410   0.022795   73.76   <2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial household taken to be 1)

    Null deviance: 479158  on 363843  levels of freedom
Residual deviance: 431627  on 363841  levels of freedom
  (17 observations deleted attributable to missingness)
AIC: 431633

Variety of Fisher Scoring iterations: 3

Let’s calculate the accuracy on our validation set.

pred <- predict(logistic_regression, df_model[val_sample,], sort = "response")
pred <- pred > imply(df_model$is_duplicate[-val_sample])
accuracy <- desk(pred, df_model$is_duplicate[val_sample]) %>% 
  prop.desk() %>% 
  diag() %>% 
  sum()
accuracy

[1] 0.6573577

We bought an accuracy of 65.7%. Not all that significantly better than random guessing.
Now let’s create our mannequin in Keras.

Mannequin definition

We are going to use a Siamese community to foretell whether or not the pairs are duplicated or not.
The thought is to create a mannequin that may embed the questions (sequence of phrases)
right into a vector. Then we will evaluate the vectors for every query utilizing a similarity
measure and inform if the questions are duplicated or not.

First let’s outline the inputs for the mannequin.

input1 <- layer_input(form = c(20), identify = "input_question1")
input2 <- layer_input(form = c(20), identify = "input_question2")

Then let’s the outline the a part of the mannequin that can embed the questions in a
vector.

word_embedder <- layer_embedding( 
  input_dim = 50000 + 2, # vocab dimension + UNK token + padding worth
  output_dim = 128,      # hyperparameter - embedding dimension
  input_length = 20,     # padding dimension,
  embeddings_regularizer = regularizer_l2(0.0001) # hyperparameter - regularization 
)

seq_embedder <- layer_lstm(
  models = 128, # hyperparameter -- sequence embedding dimension
  kernel_regularizer = regularizer_l2(0.0001) # hyperparameter - regularization 
)

Now we are going to outline the connection between the enter vectors and the embeddings
layers. Word that we use the identical layers and weights on each inputs. That’s why
that is referred to as a Siamese community. It is smart, as a result of we don’t need to have completely different
outputs if question1 is switched with question2.

vector1 <- input1 %>% word_embedder() %>% seq_embedder()
vector2 <- input2 %>% word_embedder() %>% seq_embedder()

We then outline the similarity measure we need to optimize. We wish duplicated questions
to have greater values of similarity. On this instance we’ll use the cosine similarity,
however any similarity measure might be used. Do not forget that the cosine similarity is the
normalized dot product of the vectors, however for coaching it’s not essential to
normalize the outcomes.

cosine_similarity <- layer_dot(record(vector1, vector2), axes = 1)

Subsequent, we outline a ultimate sigmoid layer to output the chance of each questions
being duplicated.

output <- cosine_similarity %>% 
  layer_dense(models = 1, activation = "sigmoid")

Now that permit’s outline the Keras mannequin when it comes to it’s inputs and outputs and
compile it. Within the compilation part we outline our loss perform and optimizer.
Like within the Kaggle problem, we are going to decrease the logloss (equal
to minimizing the binary crossentropy). We are going to use the Adam optimizer.

mannequin <- keras_model(record(input1, input2), output)
mannequin %>% compile(
  optimizer = "adam", 
  metrics = record(acc = metric_binary_accuracy), 
  loss = "binary_crossentropy"
)

We will then check out out mannequin with the abstract perform.

_______________________________________________________________________________________
Layer (sort)                Output Form       Param #    Related to                 
=======================================================================================
input_question1 (InputLayer (None, 20)         0                                       
_______________________________________________________________________________________
input_question2 (InputLayer (None, 20)         0                                       
_______________________________________________________________________________________
embedding_1 (Embedding)     (None, 20, 128)    6400256    input_question1[0][0]        
                                                          input_question2[0][0]        
_______________________________________________________________________________________
lstm_1 (LSTM)               (None, 128)        131584     embedding_1[0][0]            
                                                          embedding_1[1][0]            
_______________________________________________________________________________________
dot_1 (Dot)                 (None, 1)          0          lstm_1[0][0]                 
                                                          lstm_1[1][0]                 
_______________________________________________________________________________________
dense_1 (Dense)             (None, 1)          2          dot_1[0][0]                  
=======================================================================================
Whole params: 6,531,842
Trainable params: 6,531,842
Non-trainable params: 0
_______________________________________________________________________________________

Mannequin becoming

Now we are going to match and tune our mannequin. Nonetheless earlier than continuing let’s take a pattern for validation.

set.seed(1817328)
val_sample <- pattern.int(nrow(question1_padded), dimension = 0.1*nrow(question1_padded))

train_question1_padded <- question1_padded[-val_sample,]
train_question2_padded <- question2_padded[-val_sample,]
train_is_duplicate <- df$is_duplicate[-val_sample]

val_question1_padded <- question1_padded[val_sample,]
val_question2_padded <- question2_padded[val_sample,]
val_is_duplicate <- df$is_duplicate[val_sample]

Now we use the match() perform to coach the mannequin:

mannequin %>% match(
  record(train_question1_padded, train_question2_padded),
  train_is_duplicate, 
  batch_size = 64, 
  epochs = 10, 
  validation_data = record(
    record(val_question1_padded, val_question2_padded), 
    val_is_duplicate
  )
)

Prepare on 363861 samples, validate on 40429 samples
Epoch 1/10
363861/363861 [==============================] - 89s 245us/step - loss: 0.5860 - acc: 0.7248 - val_loss: 0.5590 - val_acc: 0.7449
Epoch 2/10
363861/363861 [==============================] - 88s 243us/step - loss: 0.5528 - acc: 0.7461 - val_loss: 0.5472 - val_acc: 0.7510
Epoch 3/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5428 - acc: 0.7536 - val_loss: 0.5439 - val_acc: 0.7515
Epoch 4/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5353 - acc: 0.7595 - val_loss: 0.5358 - val_acc: 0.7590
Epoch 5/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5299 - acc: 0.7633 - val_loss: 0.5358 - val_acc: 0.7592
Epoch 6/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5256 - acc: 0.7662 - val_loss: 0.5309 - val_acc: 0.7631
Epoch 7/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5211 - acc: 0.7701 - val_loss: 0.5349 - val_acc: 0.7586
Epoch 8/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5173 - acc: 0.7733 - val_loss: 0.5278 - val_acc: 0.7667
Epoch 9/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5138 - acc: 0.7762 - val_loss: 0.5292 - val_acc: 0.7667
Epoch 10/10
363861/363861 [==============================] - 88s 242us/step - loss: 0.5092 - acc: 0.7794 - val_loss: 0.5313 - val_acc: 0.7654

After coaching completes, we will save our mannequin for inference with the save_model_hdf5()
perform.

save_model_hdf5(mannequin, "model-question-pairs.hdf5")

Mannequin tuning

Now that now we have an affordable mannequin, let’s tune the hyperparameters utilizing the
tfruns bundle. We’ll start by including FLAGS declarations to our script for all hyperparameters we need to tune (FLAGS enable us to fluctuate hyperparmaeters with out altering our supply code):

FLAGS <- flags(
  flag_integer("vocab_size", 50000),
  flag_integer("max_len_padding", 20),
  flag_integer("embedding_size", 256),
  flag_numeric("regularization", 0.0001),
  flag_integer("seq_embedding_size", 512)
)

With this FLAGS definition we will now write our code when it comes to the flags. For instance:

input1 <- layer_input(form = c(FLAGS$max_len_padding))
input2 <- layer_input(form = c(FLAGS$max_len_padding))

embedding <- layer_embedding(
  input_dim = FLAGS$vocab_size + 2, 
  output_dim = FLAGS$embedding_size, 
  input_length = FLAGS$max_len_padding, 
  embeddings_regularizer = regularizer_l2(l = FLAGS$regularization)
)

The total supply code of the script with FLAGS might be discovered right here.

We moreover added an early stopping callback within the coaching step to be able to cease coaching
if validation loss doesn’t lower for five epochs in a row. It will hopefully cut back coaching time for dangerous fashions. We additionally added a studying fee reducer to cut back the educational fee by an element of 10 when the loss doesn’t lower for 3 epochs (this method sometimes will increase mannequin accuracy).

mannequin %>% match(
  ...,
  callbacks = record(
    callback_early_stopping(endurance = 5),
    callback_reduce_lr_on_plateau(endurance = 3)
  )
)

We will now execute a tuning run to probe for the optimum mixture of hyperparameters. We name the tuning_run() perform, passing an inventory with
the doable values for every flag. The tuning_run() perform can be chargeable for executing the script for all mixtures of hyperparameters. We additionally specify
the pattern parameter to coach the mannequin for less than a random pattern from all mixtures (lowering coaching time considerably).

library(tfruns)

runs <- tuning_run(
  "question-pairs.R", 
  flags = record(
    vocab_size = c(30000, 40000, 50000, 60000),
    max_len_padding = c(15, 20, 25),
    embedding_size = c(64, 128, 256),
    regularization = c(0.00001, 0.0001, 0.001),
    seq_embedding_size = c(128, 256, 512)
  ), 
  runs_dir = "tuning", 
  pattern = 0.2
)

The tuning run will return a knowledge.body with outcomes for all runs.
We discovered that the most effective run attained 84.9% accuracy utilizing the mix of hyperparameters proven under, so we modify our coaching script to make use of these values because the defaults:

FLAGS <- flags(
  flag_integer("vocab_size", 50000),
  flag_integer("max_len_padding", 20),
  flag_integer("embedding_size", 256),
  flag_numeric("regularization", 1e-4),
  flag_integer("seq_embedding_size", 512)
)

Making predictions

Now that now we have educated and tuned our mannequin we will begin making predictions.
At prediction time we are going to load each the textual content tokenizer and the mannequin we saved
to disk earlier.

library(keras)
mannequin <- load_model_hdf5("model-question-pairs.hdf5", compile = FALSE)
tokenizer <- load_text_tokenizer("tokenizer-question-pairs")

Since we received’t proceed coaching the mannequin, we specified the compile = FALSE argument.

Now let`s outline a perform to create predictions. On this perform we we preprocess the enter knowledge in the identical approach we preprocessed the coaching knowledge:

predict_question_pairs <- perform(mannequin, tokenizer, q1, q2) {
  q1 <- texts_to_sequences(tokenizer, record(q1))
  q2 <- texts_to_sequences(tokenizer, record(q2))
  
  q1 <- pad_sequences(q1, 20)
  q2 <- pad_sequences(q2, 20)
  
  as.numeric(predict(mannequin, record(q1, q2)))
}

We will now name it with new pairs of questions, for instance:

predict_question_pairs(
  mannequin,
  tokenizer,
  "What's R programming?",
  "What's R in programming?"
)

[1] 0.9784008

Prediction is kind of quick (~40 milliseconds).

Deploying the mannequin

To reveal deployment of the educated mannequin, we created a easy Shiny software, the place
you may paste 2 questions from Quora and discover the chance of them being duplicated. Strive altering the questions under or coming into two totally completely different questions.

The shiny software might be discovered at https://jjallaire.shinyapps.io/shiny-quora/ and it’s supply code at https://github.com/dfalbel/shiny-quora-question-pairs.

Word that when deploying a Keras mannequin you solely must load the beforehand saved mannequin file and tokenizer (no coaching knowledge or mannequin coaching steps are required).

Wrapping up

• We educated a Siamese LSTM that offers us cheap accuracy (84%). Quora’s state-of-the-art is 87%.
• We will enhance our mannequin through the use of pre-trained phrase embeddings on bigger datasets. For instance, attempt utilizing what’s described in this instance. Quora makes use of their very own full corpus to coach the phrase embeddings.
• After coaching we deployed our mannequin as a Shiny software which given two Quora questions calculates the chance of their being duplicates.