State-of-the-Art Review of Deep Learning Techniques in

Recommendation System

Ishwari Singh Rajput

Rakshita Mall

Anand Shanker Tewari

Arvind Kumar Tiwari

National Institute of Technology, Patna, Bihar, India

Kamla Nehru Institute of Technology, Sultanpur, Uttar Pradesh, India

Keywords: Deep neural network, Recommendation system, Autoencoder etc.

Abstract: On the basis of achievements of deep learning, its use in the field of recommendation system has gained a lot

of attention. This paper provides a review on the various techniques of deep learning which are used in

recommendation system. The ability of deep learning models to analyse the user-item relationship efficiently

is the reason for using it instead of the traditional recommendation models. This paper consists of the

introduction of basic terms and concepts in both recommendation system and deep learning. Then there is a

description of the researches on deep neural network based recommendation system

1 INTRODUCTION

The advancement in the field of information

technology in recent times has made the access of

huge amount of data very easy. There are a vast

variety of products and services whose description is

easily available along with their comments and

reviews. Due to this overload of information (Gantz

et al., 2012), it becomes very difficult for the user to

choose an appropriate product according to his

requirements.

To deal with the above problems,

recommendation systems are used. Recommendation

Systems provide the users with personalized

recommendations. There are many areas where

recommendation system is being used, such as music,

books, movies, shopping etc. Most of the online

vendors have a recommendation engine already

equipped. Recommendations are done on the basis of

user’s previous items choice or the items preferred by

similar user or on the basis of the description of item.

Here the item refers to the product or services which

is to be recommended. Recommendation system is

classified into three parts based on their approach:

(Çano et al., n.d.) Content based, collaborative

filtering and hybrid.

Recently, the use of artificial neural networks has

become very popular in the problems which require

complex computations and huge amount of input

data. Deep learning is a part of ANN architectural

models of deep neural network are efficiently built

and trained. Deep neural networks have its

applications in various fields such as speech

recognition, image processing, object recognition,

image processing, NLP tasks etc. Due to various

advantages of deep learning, researchers have been

encouraged to use its associated techniques in the

field of recommendation system also.

Deep learning is being used successfully in

recommendation system as well as many other fields

in computer science and has shown significant

improvements in the existing models. In 2007, a

collaborative filtering method for movie

recommendation system was given by

(Salakhutdinov et al., 2007) which utilized the

hierarchical model of deep learning.

In 2015, (Sedhain et al., 2015) with the use of auto

encoders, predicted the values which were missing in

the user-item matrix.

The sparsity issue in recommendation system in

collaborative filtering was addressed by (H. Wang et

al., 2015).

Various surveys have been done in the area of

deep neural network based recommendation system.

The state-of-the-art survey for deep recommendation

system has been done by (September & 2016, 2016).

In 2017, a comprehensive review on deep learning

based recommendation system has been done by

(Zhang et al., 2019). The paper proposed the

classifications on the basis of their structure, that is

neural network models and integration models.

182

Rajput, I., Mall, R., Tiwari, A. and Tiwari, A.

State-of-the-Art Review of Deep Learning Techniques in Recommendation System.

DOI: 10.5220/0010564900003161

In Proceedings of the 3rd International Conference on Advanced Computing and Software Engineering (ICACSE 2021), pages 182-189

ISBN: 978-989-758-544-9

This paper is organized in various sections as

follows. First, it consists of the introduction of

recommendation and deep learning techniques.

Section two consists of background and related

terminologies. In Section three, there is a review of

the various approaches of deep neural networks in the

area of recommendation system and its classification.

Section 5 includes the concluding part.

2 BACKGROUND AND

TERMINOLOGIES

A recommendation system is used for information

filtering and outputs a list of specific products in a

personalized manner. For examining how the two

fields that is recommendation system and deep

learning are integrated together, one should know the

basic of both these fields. This section of the paper

includes a brief description about the fundamental

classifications and challenges of both the fields. First,

there is an introduction of types of recommendation

system and then the details about deep learning

methods.

2.1 Recommendation System

It is very important for a user to filter the huge amount

of data available, to find a useful, tailored and

relevant content. The recommendations that are

predicted by the recommendation System helps the

users in taking a decision.

In a traditional recommendation system, the

recommendations can be made in two different

manners, i.e., predicting a particular item or preparing

a ranking list of items for a particular user (Park et al.,

n.d.). The recommendation system is divided into

three broad categories: Content based (Park et al.,

n.d.), Collaborative filtering and hybrid

recommendation system (Park et al., n.d.).

 Content based recommendation system: In

content-based recommendation, the items which

are similar in content is searched. The profile of

user is established on the basis of items on which

the user is interested in. According to the profile

generated, the recommendation system searches

the database for the appropriate items using the

descriptive attributes of the item. If we use this

recommendation system (Lops et al., 2011) for

an item which is newly added, then content based

recommendation system works very efficiently.

The problem with new inserted item is that it may

not have any rating, but still the algorithm works

since it uses descriptive information for

recommendations. The limitation of this method

is that it cannot recommend diverse range of

products since the algorithm does not take the

information from similar users.

 Collaborative filtering recommendation systems:

This recommendation system assumes that users’

who have previously preferred same items,

would have same choice in future also. In this

system, the recommendations are done on the

basis of similar users’ pattern rather than

descriptive features of items. A correlation

among the users is determined, depending upon

the choice of similar users, the items are

recommended.

There are two methods which the CF algorithm

follows: memory based algorithm and model

based algorithms.

In memory based algorithm the complete user-

item matrix is taken into consideration for

identifying similarity. After finding out the

nearest neighbour, on the basis of neighbours

past rating, the recommendations are provided.

In model based algorithms, an offline model is

built with the use of machine learning algorithms

and data mining methods. These models include

clustering models, Decision models, Bayesian

model and singular value decomposition model.

 Hybrid Recommendation System: Hybrid

recommendation system is a combination of

content based and collaborative filtering model,

it incorporates the benefits of both the methods

and tries to eliminate the limitations of the above

models (Tran et al., 2000). There are many

hybrid systems proposed, some of them are as

following:

Cascade: The output of one approach is later on

used by other approach

Switching: Recommendation output is given on

the basis of current situation and either of the one

approach is used.

Weighted: the combination of the scores of

various approaches is used for recommendation

2.2 Deep Learning

Deep learning is based on learning many layers of

representations with the help of artificial neural

networks, and is a part of machine learning. Deep

learning has its applications in various fields such as

Computer vision, recognition of speech, natural

language processing etc. The important factors which

increase the importance of deep neural network as the

State-of-the-Art Review of Deep Learning Techniques in Recommendation System

183

state-of-the-art machine learning methods are as

following:

Big data: As the amount of data increases, better

representations are learnt by the deep learning model.

Computational power: The complex computations of

the deep learning model is done using the GPU.

In this section, there is a description of various deep

learning models which are used in recommendation

systems.

2.2.1 Autoencoder

It is a type of a feed forward network in which some

representations from the encoded input are found by

the training, so that the input can be restored back

from these representations. An autoencoder consists

of three layers, which are the input layer, the hidden

layer and the output layer. There are equivalent

numbers of neurons in the input layer and the output

layer. The representations are obtained from the

hidden layer, and with the help of these

representations, the input layer is reconstructed at the

output layer (Deng et al., n.d.).

In the learning process, two mappings are used,

with the help of encoder and decoder. Encoder is used

for mapping the data form input to hidden layer while

decoder is used to map the data from hidden to output

layer (Strub et al., n.d.).

2.2.2 Recurrent Neural Network

The use of sequential information is done in RNN,

which is a class of artificial neural network. In RNN,

the sequence of values, i.e., x

(0)

, x

(1)

,..., x

(t)

processed. On each element of the sequence, the same

task is performed and the output is based on previous

computations. RNN (Wu et al., n.d.) uses an internal

memory to hold the values of previous computations,

so that it may be used later. Some major problems that

occur in RNN is exploding gradient or vanishing

gradient. To remove this problem, Long short term

memory (LSTM) and gated recurrent unit is used. It

is used when the processing is to be done for

predicting events which has comparatively longer

interval and delays. LSTM has a processor to

distinguish the useful information from the

information which are not useful. This processor is

known as cell. There are three gates in LSTM namely

input gate, output gate and forget gate. The

information that does not complies with the

algorithm’s certification, is forgotten with the help of

forgot gate. RNN are useful while dealing with

temporal dynamics of interactions and when the

user’s behavior has a certain sequence of pattern.

2.2.3 Convolution Neural Network

It is a neural network of fed forward type. In CNN

(Wu et al., n.d.), the use of convolution operations is

done rather than the usual matrix multiplication in

one or more layers. There are many applications in

which CNNs are applied such as object recognition,

self-driving cars, audio processing etc. while

transforming the input to output, the CNN uses three

major components that are convolution layer, pooling

layer and fully connected layers. These three are

stacked together to form a CNN. The layers of CNN

are used for the following operations.

Convolution: It is a core operation and is used for

extracting the features from the input. It is done by

applying convolution filters involving some

mathematical operation.

Non linearity: An additional operation is used for

introducing non linearity after every convolution

operation and usually ReLU is used.

Pooling: This operation is done for reducing the

dimensionality of the feature maps and decreases the

processing time.

Figure 1: Convolutional Neural Network Model

2.2.4 Restricted Boltzmann Machine

It consists of two layers of neural network namely a

visible and a hidden layer. The complex computations

and learning in RBM (Wu et al., n.d.) is based on

inherent intrinsic expression of data. The word

“restricted’ is used for intra-layer communication as

it is not present in both hidden layer and visible layer.

Due to this restriction, the learning efficiency

increases. There is a full connection between the

nodes of different layers which are stacked together

and there is no connection between the nodes of same

layer. Since RBM uses a simple forward encoding

operation, so it is very fast when compared to other

models such as autoencoder.

ICACSE 2021 - International Conference on Advanced Computing and Software Engineering

184

3 DEEP NEURAL NETWORK

BASED RECOMMENDATION

SYSTEM

DL techniques are widely used in various real world

applications such as sentiment analysis, speech

recognition image classification, text classification

etc. Various researchers also include deep neural

network based techniques in the field of

recommendation system to improve its performance

as compared to traditional recommendations systems.

This section describes the various categories

recommendation systems which are based on deep

learning. The categorization is based on the types of

recommendation systems used which is as follows:

 Collaborative filtering recommendation system

based on deep neural network.

 Content-based recommendation system based on

deep neural network.

 Hybrid recommendation system based on deep

neural network.

 Social network-based recommendation system

based on deep neural network.

 Context aware recommendation system based on

deep neural network.

Integration model and neural network model are

the two categories of deep neural network based

recommendation system. Integration model is further

divided into two categories on the basis of whether it

combines any traditional recommendation system

model with deep neural network technique or depends

solely on deep learning method.

Neural network model is also divided into two

categories on the basis of deep neural network based

technique used: models which uses single deep neural

network based technique and deep

neural network

based composite model. In deep neural network based

composite model, different deep neural network

techniques are used to build a hybrid system having

more capability. The framework of recommendation

system

based on deep neural network is presented in

Figure 2.

Figure 2. Deep neural network based Recommendation

system

3.1 Collaborative Filtering

Recommendation Systems based on

Deep Neural Networks

Collaborative filtering (CF) is one of the commonly

implemented techniques in recommendation systems

in order to tackle various real-life issues. The state of

the art CF-based methods uses the rating matrix for

recommending the items. But this approach faces the

problem of data sparseness and cold start problem.

The sparsity of the user-item matrix, the learned

features is not effective which reduces the

performance of recommendation system. Various

researchers propose deep neural network based

collaborative filtering techniques to enhance its

effectiveness in recommendation.

3.1.1 Collaborative Filtering Method based

on Generative Adversarial Network

Generative Adversarial Network is a neural network

which is generative and having discriminator and

generator functions. These both functions are

simultaneously trained in competition with one

another in architecture of minimax game. The first

model to implement GAN in the field of Information

Retrieval is (IRGAN) (J. Wang et al., 2017) which

stands for Information retrieval generative

adversarial network. The state of the art GAN model

has two modules a discriminator and a generator. The

generative retrieval module predicts appropriate

documents with given query, whereas discriminative

retrieval module predicts relevancy given with a pair

of query and document.

The IRGAN model combines above two

Information Retrieval models in order to play a

minimax game with them: the generative retrieval

model produces (or selects) relevant documents that

are relevant documents like ground truth, while the

discriminating retrieval model separates the relevant

documents from those generated by the generative

retrieval model.

3.1.2 Recurrent Neural Network based

Collaborative Filtering Method

In order to deal with the information in sequential

form, recurrent neural network (RNN) proves to be a

very effective network. Concepts of loops are used in

place of feedforward network to remember

sequences. Variants of RNN viz. Long Short-Term

Memory (LSTM) and Gated Recurrent Unit (GRU)

network are used to deal with the problems of long

term dependencies and vanishing gradient problem.

State-of-the-Art Review of Deep Learning Techniques in Recommendation System

185

In collaborative filtering method based on RNN, the

impact of user historical sequence is modelled on the

current behavior of user, recommendation is

performed and user’s behavior is predicted [19].

Figure 7 shows the framework of collaborative

filtering method based on RNN (Wu et al., n.d.). Let

the input set is {I

, I

. . . I

}, and output is O

= σ (f

(Wꞏh

t−1

+ VꞏI

)ꞏV), σ represents a softmax function, f

represents the activation function, which specifies the

selection probability of any item at time t. h

represents the hidden state vector.

Figure 3. Collaborative filtering model based on RNN.

3.1.3 Collaborative Filtering Method based

on Autoencoder

The first ever developed autoencoder-based

collaborative recommendation model is Autoencoder

based Collaborative filtering (Sedhain et al., 2015). It

decomposes the vectors by integer ratings. The model

proposed by (Sedhain et al., 2015) takes user or item

based ratings as inputs in rating matrix R. The output

is produced by the process of encoding and decoding

by optimizing the parameters of model and reducing

the reconstruction error. Consider an example, if the

range of integers [1-5] represents the rating score,

then each r

can be divided into five vectors.

Figure 4. Collaborative filtering method based on

Autoencoder

Above figure represents the 1 to 5 rating scale in

which blue boxes represents the user rated item. The

cost function which is to be reduced is taken as Mean

Square Error. The rating prediction in this approach

is found by making the summary of each of the five

vectors, which are scaled by rating K. Pretraining of

parameters and local optimum avoidance is

performed by RBM. Stacking multiple autoencoder

collectively shows the slight improvement in

accuracy. This method based on autoencoder suffers

from the problem of dealing with non-integer ratings

and sparseness of input data due to decomposition of

partial observed vectors.

Collaborative Denoising Auto-Encoder (Strub et

al., n.d.) is primarily used for prediction rankings.

User feedback is taken as input to the CDAE. If the

user enjoys a movie, the input value is 1 otherwise it

is 0. It shows the vector preference to display the

user's interest in some item. Gaussian noise corrupts

the CDAE input.

3.1.4 Collaborative Filtering Method based

on Restricted Boltzmann Machine

Restricted Boltzmann Machine (RBM) is a two-layer

neural network capable to deal with typical learning

based problems. The efficiency of learning is

improved by removing the connections between same

layers. This recommendation method is proposed by

(Salakhutdinov et al., 2007). Further a conditional

RBM model is proposed to consider information in

form of feedback. The visible layer of RBM can take

only binary values so only one hot vector can be used

to represent the rating score. The architecture of RBM

based model is represented as shown in Figure 5.

Figure 5. Collaborative filtering model based on RBM.

There are equal hidden layers in each RBM and

have softmax units which are visible for movie

ratings given by user. There is one training case in

ICACSE 2021 - International Conference on Advanced Computing and Software Engineering

186

each RBM unit with combined weights and biases.

Hidden units hold the binary states which are

different for separate users. Let there are N movies

rated by user and there are n visible units. Suppose M

represents X × N size matrix where 𝑟





= 1 if movie i

is rated as x by user u, otherwise 𝑟





= 0. So,

𝑝



𝑟





1

ℎ





 







∑















∑

 













∑

















(1)

𝑝ℎ



 1M  σ b





∑∑

𝑟

















) (2)

where σ (x) represents a logistic function, 𝑟





interaction, 𝑏





:bias of rating x for movie i and b

hidden unit bias.

3.2 Content-based Recommendation

Systems based on Deep Neural

Networks

Content-based recommendation systems recommend

on the basis of descriptive attributes of items and

users’ profiles such as texts, pictures and videos

(Meteren et al., n.d.).

The use of deep neural network in content-based

recommendation systems is to capture the non-linear

relationships between user and item (Meteren et al.,

n.d.). It also captures the intra-relationship between

data, from large available data sources.

3.3 Hybrid Recommendation System

based on Deep Neural Networks

The state-of-the-art CF-based approach uses the

ranking matrix to suggest products. But this method

suffers from a data sparsity and a cold start crisis. Due

to the scarce existence of the user-item matrix, the

features learned are not accurate, which decreases the

efficiency of the recommendation framework. Hybrid

recommendation model based on deep neural

networks incorporates a content-based

recommendation model with collective

recommendation-based filtering models, which

combines the mechanism of feature learning and

recommendation into a single model. (Meteren et al.,

n.d.) suggested a layered, self-encoder-based hybrid

paradigm that learns the latent space factors of users

and items and simultaneously performs mutual

filtering from the ranking matrix.

An autoencoder is a variant of neural network,

having encoder and decoder as two components. The

encoder converts the input into its hidden

representation, while the decoder converts the hidden

representation back to the restored input form. The

parameters corresponding to the autoencoder are

trained to minimize the error due to the

reconstruction, which is measured by the loss

function. Denoising autoencoder (DAE) tries to

reconstruct the input from a corrupted version for

improved representation from the input. More

variants of autoencoder have been developed for

better outcomes. The hybrid recommendation model

based on the stacked denoising autoencoder uses both

the rating matrix and the side information and

integrates both the SDAE (Meteren et al., n.d.) and

the matrix factorisation. Matrix factorization is a

widely used model with improved accuracy, and

SDAE is a powerful model for extracting high-level

features from inputs. The integration of the above two

model will become a powerful model for more

benefits.

3.4 Social Network-based

Recommendation System using

Deep Neural Networks

Conventional recommendation models never

consider social connections among the users. But we

always take verbal recommendations from our friends

in reality. These verbal recommendations are termed

as social recommendation which occurs daily

(Meteren et al., n.d.). Hence, for improved

recommendation systems and for more personalized

recommendations, social network must be employed

among users. Every user will interact with various

types of social relationships.

The quality of recommendation system is very

crucial task which can be achieved by implementing

the effect of social relationship among the users.

Items with location attributes and sequential pattern

of user behaviour in spatial and temporal frame are

used to form spatiotemporal pattern which is used to

improve recommendation accuracy. Recently, a very

few recommendation techniques have been proposed

which is based on the users’ trust relations improve

conventional recommendation systems. These trust

based recommendation models proves to be an

effective move in the field of recommendation system

models.

In current scenario an integration of deep learning

and social network based recommendation system

provides a platform for various research solutions.

The limitations which are inherent to the social

recommendation must be addressed in the future

research.

State-of-the-Art Review of Deep Learning Techniques in Recommendation System

187

3.5 Context-aware Recommendation

Systems based on Deep Neural

Networks

A context-aware recommendation system, integrates

context based information into a recommendation

model. This integration is effectively performed by

the deep learning techniques in different conditions of

recommending item. Deep neural network based

methods are used to extract the latent space

presentation from the context based information.

Deep learning-based model can be integrated into

diverse data to reduce data sparsity problem.

Sequential nature of data plays a significant part in

implementing user behaviours. Recently, recurrent

neural networks (RNNs) are commonly used in a

variety of sequential simulation activities. However,

for real-world implementations, these approaches

have trouble modeling contextual knowledge, which

has been shown to be very essential for behavioural

modelling.

Currently this method based on deep neural

networks focused towards situation information. A

novel approach is proposed called context-aware

recurrent neural networks. It uses two types of

matrices: input and transition matrices. They both are

specific to the context and adaptive in nature. Input

matrices are used to extract various situations such as

time, place, weather condition where actually the user

behaves.

3.6 Comparisons

As deep learning plays a significant role in most of

the fields as it has the capability of dealing with large

and complex problems with improved results. Deep

learning technology also contributes in the field of

recommendation system for improved customer

satisfaction. Deep learning technology overcomes the

shortcomings of traditional models to get high quality

recommendations (Liang et al., 2018).

All the above discussed methods of

recommendation systems use deep neural networks

and hence also achieve the quality in recommending

items to the users (Liang et al., 2018). Different

recommendation models use different deep learning

methods to obtain improve results.

The summary of various deep neural network

based recommendation systems are given below:

Table 1: Comparison

Application

of Deep

Learning in

Findin

Conten

ased

recommendation

stems

Extract non-linear use

item relationship and intricate

relationship with data itself.

Collaborative

filtering

recommendation

systems

Takes the use

-item

ranking matrix as input and

uses a deep neural network

based model to learn the latent

space presentation that

corresponds to users or

objects. Use the loss function

to construct a deep neural

network-based model

optimization function. On the

basis of the latent space

presentation,

recommendations are made.

Hybrid

recommendation

systems

Integrate the individual or

object learning process and the

suggestion process into a

single architecture.

Social network-

based

recommendation

systems

Focuses on social

relationship between users,

and extracts the effect of

location of user, movement

patterns and other various

factors.

Context-aware

recommendation

systems

Deep neural network

based methods combines the

context information into the

recommendation model and

obtain the latent space

presentation of the context

based information. Also

reduce data sparsit

in model.

4 CONCLUSIONS

The massive volume of data generates the necessity

of applications and technology to efficiently process

and information analysis for providing the maximum

benefit to the target users. Recommendation system

based on deep neural networks proves to be optimal

solution for above mentioned challenges. Deep neural

network based recommendation systems can easily

learn the features of items along with users from large

volume of information, to build a system for effective

recommendation to users. This article compares the

conventional recommendation models with the deep

neural network based recommendation techniques

which extracts the features of users along with items

ICACSE 2021 - International Conference on Advanced Computing and Software Engineering

188

to increase the accuracy of recommending items to

users. It also integrates data from multiple sources to

extract the preferences of users.

REFERENCES

Çano, E., Analysis, M. M.-I. D., & 2017, undefined. (n.d.).

Hybrid recommender systems: A systematic literature

review. Content.Iospress.Com. Retrieved January 19,

2022, from

https://content.iospress.com/articles/intelligent-data-

analysis/ida163209

Deng, S., Huang, L., Xu, G., … X. W.-I. transactions on, &

2016, undefined. (n.d.). On deep learning for trust-

aware recommendations in social networks.

Ieeexplore.Ieee.Org. Retrieved January 19, 2022, from

https://ieeexplore.ieee.org/abstract/document/7414528/

Gantz, J., future, D. R.-I. iView: I. A. the, & 2012,

undefined. (2012). The digital universe in 2020: Big

data, bigger digital shadows, and biggest growth in the

far east. Speicherguide.De.

https://www.speicherguide.de/download/dokus/IDC-

Digital-Universe-Studie-iView-11.12.pdf

Liang, N., Zheng, H. T., Chen, J. Y., Sangaiah, A. K., &

Zhao, C. Z. (2018). TRSDL: Tag-Aware Recommender

System Based on Deep Learning–Intelligent

Computing Systems. Applied Sciences 2018, Vol. 8,

Page 799, 8(5), 799.

https://doi.org/10.3390/APP8050799

Lops, P., Gemmis, M. De, handbook, G. S.-R. systems, &

2011, undefined. (2011). Content-based recommender

systems: State of the art and trends. Springer, 73–105.

https://doi.org/10.1007/978-0-387-85820-3_3

Meteren, R. Van, … M. V. S. in the new information, &

2000, undefined. (n.d.). Using content-based filtering

for recommendation. Users.Ics.Forth.Gr. Retrieved

January 19, 2022, from

http://users.ics.forth.gr/~potamias/mlnia/paper_6.pdf

Park, D., Kim, H., Choi, I., applications, J. K.-E. systems

with, & 2012, undefined. (n.d.). A literature review and

classification of recommender systems research.

Elsevier. Retrieved January 19, 2022, from

https://www.sciencedirect.com/science/article/pii/S095

7417412002825

Salakhutdinov, R., Mnih, A., & Hinton, G. (2007).

Restricted Boltzmann machines for collaborative

filtering. ACM International Conference Proceeding

Series, 227, 791–798.

https://doi.org/10.1145/1273496.1273596

Sedhain, S., Menon, A., Sanner, S., … L. X.-24th

international conference, & 2015, undefined. (2015).

Autorec: Autoencoders meet collaborative filtering.

Dl.Acm.Org, 111–112.

https://doi.org/10.1145/2740908.2742726

September, L. Z., & 2016, undefined. (2016). A survey and

critique of deep learning on recommender systems.

Bdsc.Lab.Uic.Edu.

https://bdsc.lab.uic.edu/docs/survey-critique-deep.pdf

Strub, F., eCommerce, J. M.-N. workshop on machine

learning for, & 2015, undefined. (n.d.). Collaborative

filtering with stacked denoising autoencoders and

sparse inputs. Hal.Inria.Fr. Retrieved January 19,

2022, from https://hal.inria.fr/hal-01256422/

Tran, T., Markets, R. C.-P. K.-B. E., Papers, undefined, &

2000, undefined. (2000). Hybrid recommender systems

for electronic commerce. Aaai.Org.

https://www.aaai.org/Papers/Workshops/2000/WS-00-

04/WS00-04-012.pdf

Wang, H., Wang, N., … D. Y.-A. S. international

conference, & 2015, undefined. (2015). Collaborative

deep learning for recommender systems. Dl.Acm.Org,

2015-August, 1235–1244.

https://doi.org/10.1145/2783258.2783273

Wang, J., Yu, L., Zhang, W., Gong, Y., Xu, Y., Wang, B.,

Zhang, P., & Zhang, D. (2017). IRGAN: A minimax

game for unifying generative and discriminative

information retrieval models. SIGIR 2017 -

Proceedings of the 40th International ACM SIGIR

Conference on Research and Development in

Information Retrieval, 515–524.

https://doi.org/10.1145/3077136.3080786

Wu, C., Wang, J., Liu, J., Systems, W. L.-K.-B., & 2016,

undefined. (n.d.). Recurrent neural network based

recommendation for time heterogeneous feedback.

Elsevier. Retrieved January 19, 2022, from

https://www.sciencedirect.com/science/article/pii/S095

070511630199X

Zhang, S., Yao, L., Sun, A., & Tay, Y. (2019). Deep

learning based recommender system: A survey and new

perspectives. In ACM Computing Surveys (Vol. 52,

Issue 1). Association for Computing Machinery.

https://doi.org/10.1145/3285029

State-of-the-Art Review of Deep Learning Techniques in Recommendation System

189