Incorporating LDA with LSTM for followee recommendation on Twitter network

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Brahim Dib ◽  
Fahd Kalloubi ◽  
El Habib Nfaoui ◽  
Abdelhak Boulaalam
Keyword(s):  
Neural Network ◽  
Recommendation System ◽  
Latent Dirichlet Allocation ◽  
Short Term Memory ◽  
Real Life ◽  
Network Graph ◽  
Data Set ◽  
Content Type ◽  
Modeling Strategy ◽  
Searching Method

Purpose The purpose of this study is to facilitate the task of finding appropriate information to read about, and searching for people who are in the same field of interest. Knowing that more people keep up with new streaming information on Twitter micro-blogging service. With the immense number of micro-posts shared via the follower/followee network graph, Twitter users find themselves in front of millions of tweets, which makes the task crucial. Design/methodology/approach In this paper, a long short–term memory (LSTM) model that relies on the latent Dirichlet allocation (LDA) output vector for followee recommendation, the LDA model applied as a topic modeling strategy is proposed. Findings This study trains the model using a real-life data set extracted based on Twitter follower/followee architecture. It confirms the effectiveness and scalability of the proposed approach. The approach improves the state-of-the-art models average-LSTM and time-LSTM. Research limitations/implications This study improves mainly the existing followee recommendation systems. Because, unlike previous studies, it applied a non-hand-crafted method which is the LSTM neural network with LDA model for topics extraction. The main limitation of this study is the cold-start users cannot be treated, also some active fake accounts may not be detected. Practical implications The aim of this approach is to assist users seeking appropriate information to read about, by choosing appropriate profiles to follow. Social implications This approach consolidates the social relationship between users in a microblogging platform by suggesting like-minded people to each other. Thus, finding users with the same interests will be easy without spending a lot of time seeking relevant users. Originality/value Instead of classic recommendation models, the paper provides an efficient neural network searching method to make it easier to find appropriate users to follow. Therefore, affording an effective followee recommendation system.

Comparative study on performance of different artificial neural network methods for prediction of the Covid19

foresight ◽  
2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Alireza Sedighi Fard
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Short Term Memory ◽  
Polynomial Regression ◽  
World Health ◽  
Comparative Approach ◽  
Data Set ◽  
Current Time ◽  
Content Type ◽  
Artificial Neural

Purpose This study aims to compare many artificial neural network (ANN) methods to find out which method is better for the prediction of Covid19 number of cases in N steps ahead of the current time. Therefore, the authors can be more ready for similar issues in the future. Design/methodology/approach The authors are going to use many ANNs in this study including, five different long short-term memory (LSTM) methods, polynomial regression (from degree 2 to 5) and online dynamic unsupervised feedforward neural network (ODUFFNN). The authors are going to use these networks over a data set of Covid19 number of cases gathered by World Health Organization. After 1,000 epochs for each network, the authors are going to calculate the accuracy of each network, to be able to compare these networks by their performance and choose the best method for the prediction of Covid19. Findings The authors concluded that for most of the cases LSTM could predict Covid19 cases with an accuracy of more than 85% after LSTM networks ODUFFNN had medium accuracy of 45% but this network is highly flexible and fast computing. The authors concluded that polynomial regression cant is a good method for the specific purpose. Originality/value Considering the fact that Covid19 is a new global issue, less studies have been conducted with a comparative approach toward the prediction of Covid19 using ANN methods to introduce the best model of the prediction of this virus.

Pattern-based dual learning for point-of-interest (POI) recommendation

2020 ◽  
Vol 120 (10) ◽  
pp. 1901-1921
Author(s):  
Tipajin Thaipisutikul ◽  
Yi-Cheng Chen
Keyword(s):  
Recommendation System ◽  
Short Term Memory ◽  
Real Life ◽  
Experimental Results ◽  
Short Term ◽  
Term Memory ◽  
Content Type ◽  
Point Of Interest ◽  
Poi Recommendation ◽  
Long Short Term Memory

PurposeTourism spot or point-of-interest (POI) recommendation has become a common service in people's daily life. The purpose of this paper is to model users' check-in history in order to predict a set of locations that a user may soon visit.Design/methodology/approachThe authors proposed a novel learning-based method, the pattern-based dual learning POI recommendation system as a solution to consider users' interests and the uniformity of popular POI patterns when making recommendations. Differing from traditional long short-term memory (LSTM), a new users’ regularity–POIs’ popularity patterns long short-term memory (UP-LSTM) model was developed to concurrently combine the behaviors of a specific user and common users.FindingsThe authors introduced the concept of dual learning for POI recommendation. Several performance evaluations were conducted on real-life mobility data sets to demonstrate the effectiveness and practicability of POI recommendations. The metrics such as hit rate, precision, recall and F-measure were used to measure the capability of ranking and precise prediction of the proposed model over all baselines. The experimental results indicated that the proposed UP-LSTM model consistently outperformed the state-of-the-art models in all metrics by a large margin.Originality/valueThis study contributes to the existing literature by incorporating a novel pattern–based technique to analyze how the popularity of POIs affects the next move of a particular user. Also, the authors have proposed an effective fusing scheme to boost the prediction performance in the proposed UP-LSTM model. The experimental results and discussions indicate that the combination of the user's regularity and the POIs’ popularity patterns in PDLRec could significantly enhance the performance of POI recommendation.

Comparing predictive performance of general regression neural network (GRNN) and hedonic regression model for factors affecting housing prices in “Pune-India”

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Abhijat Arun Abhyankar ◽  
Harish Kumar Singla
Keyword(s):  
Neural Network ◽  
Regression Model ◽  
Multivariate Regression ◽  
Housing Prices ◽  
Predictive Performance ◽  
General Regression Neural Network ◽  
Multivariate Regression Model ◽  
Data Set ◽  
Content Type ◽  
General Regression

Purpose The purpose of this study is to compare the predictive performance of the hedonic multivariate regression model with the probabilistic neural network (PNN)-based general regression neural network (GRNN) model of housing prices in “Pune-India.” Design/methodology/approach Data on 211 properties across “Pune city-India” is collected. The price per square feet is considered as a dependent variable whereas distances from important landmarks such as railway station, fort, university, airport, hospital, temple, parks, solid waste site and stadium are considered as independent variables along with a dummy for amenities. The data is analyzed using a hedonic type multivariate regression model and GRNN. The GRNN divides the entire data set into two sets, namely, training set and testing set and establishes a functional relationship between the dependent and target variables based on the probability density function of the training data (Alomair and Garrouch, 2016). Findings While comparing the performance of the hedonic multivariate regression model and PNN-based GRNN, the study finds that the output variable (i.e. price) has been accurately predicted by the GRNN model. All the 42 observations of the testing set are correctly classified giving an accuracy rate of 100%. According to Cortez (2015), a value close to 100% indicates that the model can correctly classify the test data set. Further, the root mean square error (RMSE) value for the final testing for the GRNN model is 0.089 compared to 0.146 for the hedonic multivariate regression model. A lesser value of RMSE indicates that the model contains smaller errors and is a better fit. Therefore, it is concluded that GRNN is a better model to predict the housing price functions. The distance from the solid waste site has the highest degree of variable senstivity impact on the housing prices (22.59%) followed by distance from university (17.78%) and fort (17.73%). Research limitations/implications The study being a “case” is restricted to a particular geographic location hence, the findings of the study cannot be generalized. Further, as the objective of the study is restricted to just to compare the predictive performance of two models, it is felt appropriate to restrict the scope of work by focusing only on “location specific hedonic factors,” as determinants of housing prices. Practical implications The study opens up a new dimension for scholars working in the field of housing prices/valuation. Authors do not rule out the use of traditional statistical techniques such as ordinary least square regression but strongly recommend that it is high time scholars use advanced statistical methods to develop the domain. The application of GRNN, artificial intelligence or other techniques such as auto regressive integrated moving average and vector auto regression modeling helps analyze the data in a much more sophisticated manner and help come up with more robust and conclusive evidence. Originality/value To the best of the author’s knowledge, it is the first case study that compares the predictive performance of the hedonic multivariate regression model with the PNN-based GRNN model for housing prices in India.

MDTP

2021 ◽  
Vol 14 (8) ◽  
pp. 1289-1297
Author(s):  
Ziquan Fang ◽  
Lu Pan ◽  
Lu Chen ◽  
Yuntao Du ◽  
Yunjun Gao
Keyword(s):  
Neural Network ◽  
Short Term Memory ◽  
Temporal Dynamics ◽  
Real Life ◽  
Feature Modeling ◽  
Traffic Prediction ◽  
Interactive System ◽  
Trajectory Data ◽  
Spatio Temporal ◽  
Prediction Approach

Traffic prediction has drawn increasing attention for its ubiquitous real-life applications in traffic management, urban computing, public safety, and so on. Recently, the availability of massive trajectory data and the success of deep learning motivate a plethora of deep traffic prediction studies. However, the existing neural-network-based approaches tend to ignore the correlations between multiple types of moving objects located in the same spatio-temporal traffic area, which is suboptimal for traffic prediction analytics. In this paper, we propose a multi-source deep traffic prediction framework over spatio-temporal trajectory data, termed as MDTP. The framework includes two phases: spatio-temporal feature modeling and multi-source bridging. We present an enhanced graph convolutional network (GCN) model combined with long short-term memory network (LSTM) to capture the spatial dependencies and temporal dynamics of traffic in the feature modeling phase. In the multi-source bridging phase, we propose two methods, Sum and Concat, to connect the learned features from different trajectory data sources. Extensive experiments on two real-life datasets show that MDTP i) has superior efficiency, compared with classical time-series methods, machine learning methods, and state-of-the-art neural-network-based approaches; ii) offers a significant performance improvement over the single-source traffic prediction approach; and iii) performs traffic predictions in seconds even on tens of millions of trajectory data. we develop MDTP + , a user-friendly interactive system to demonstrate traffic prediction analysis.

Mining numerical measure of consumers’ product evaluation expressed in words based on latent Dirichlet allocation

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Ziang Wang ◽  
Feng Yang
Keyword(s):  
Latent Dirichlet Allocation ◽  
Product Evaluation ◽  
Online Reviews ◽  
Product Evaluations ◽  
Product Attributes ◽  
Data Set ◽  
Content Type ◽  
Face To Face ◽  
Online Retailers ◽  
Dirichlet Allocation

Purpose It has always been a hot topic for online retailers to obtain consumers’ product evaluations from massive online reviews. In the process of online shopping, there is no face-to-face interaction between online retailers and customers. After collecting online reviews left by customers, online retailers are eager to acquire answers to some questions. For example, which product attributes will attract consumers? Or which step brings a better experience to consumers during the process of shopping? This paper aims to associate the latent Dirichlet allocation (LDA) model with the consumers’ attitude and provides a method to calculate the numerical measure of consumers’ product evaluation expressed in each word. Design/methodology/approach First, all possible pairs of reviews are organized as a document to build the corpus. After that, latent topics of the traditional LDA model noted as the standard LDA model, are separated into shared and differential topics. Then, the authors associate the model with consumers’ attitudes toward each review which is distinguished as positive review and non-positive review. The product evaluation reflected in consumers’ binary attitude is expanded to each word that appeared in the corpus. Finally, a variational optimization is introduced to calculate parameters mentioned in the expanded LDA model. Findings The experiment’s result illustrates that the LDA model in the research noted as an expanded LDA model, can successfully assign sufficient probability with words related to products attributes or consumers’ product evaluation. Compared with the standard LDA model, the expanded model intended to assign higher probability with words, which have a higher ranking within each topic. Besides, the expanded model also has higher precision on the prediction set, which shows that breaking down the topics into two categories fits better on the data set than the standard LDA model. The product evaluation of each word is calculated by the expanded model and depicted at the end of the experiment. Originality/value This research provides a new method to calculate consumers’ product evaluation from reviews in the level of words. Words may be used to describe product attributes or consumers’ experiences in reviews. Assigning words with numerical measures can analyze consumers’ products evaluation quantitatively. Besides, words are labeled themselves, they can also be ranked if a numerical measure is given. Online retailers can benefit from the result for label choosing, advertising or product recommendation.

Tunicate swarm algorithm-trained multi-layered perceptron for data centre energy demand forecasting and relative percentage contribution analysis of input parameters

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Oluwafemi Ajayi ◽  
Reolyn Heymann
Keyword(s):  
Neural Network ◽  
Energy Management ◽  
Energy Demand ◽  
Mean Squared Error ◽  
Data Set ◽  
Content Type ◽  
Demand Pattern ◽  
The Neural Network ◽  
Input Parameters ◽  
Demand Profile

Purpose Energy management is critical to data centres (DCs) majorly because they are high energy-consuming facilities and demand for their services continue to rise due to rapidly increasing global demand for cloud services and other technological services. This projected sectoral growth is expected to translate into increased energy demand from the sector, which is already considered a major energy consumer unless innovative steps are used to drive effective energy management systems. The purpose of this study is to provide insights into the expected energy demand of the DC and the impact each measured parameter has on the building's energy demand profile. This serves as a basis for the design of an effective energy management system. Design/methodology/approach This study proposes novel tunicate swarm algorithm (TSA) for training an artificial neural network model used for predicting the energy demand of a DC. The objective is to find the optimal weights and biases of the model while avoiding commonly faced challenges when using the backpropagation algorithm. The model implementation is based on historical energy consumption data of an anonymous DC operator in Cape Town, South Africa. The data set provided consists of variables such as ambient temperature, ambient relative humidity, chiller output temperature and computer room air conditioning air supply temperature, which serve as inputs to the neural network that is designed to predict the DC’s hourly energy consumption for July 2020. Upon preprocessing of the data set, total sample number for each represented variable was 464. The 80:20 splitting ratio was used to divide the data set into training and testing set respectively, making 452 samples for the training set and 112 samples for the testing set. A weights-based approach has also been used to analyze the relative impact of the model’s input parameters on the DC’s energy demand pattern. Findings The performance of the proposed model has been compared with those of neural network models trained using state of the art algorithms such as moth flame optimization, whale optimization algorithm and ant lion optimizer. From analysis, it was found that the proposed TSA outperformed the other methods in training the model based on their mean squared error, root mean squared error, mean absolute error, mean absolute percentage error and prediction accuracy. Analyzing the relative percentage contribution of the model's input parameters based on the weights of the neural network also shows that the ambient temperature of the DC has the highest impact on the building’s energy demand pattern. Research limitations/implications The proposed novel model can be applied to solving other complex engineering problems such as regression and classification. The methodology for optimizing the multi-layered perceptron neural network can also be further applied to other forms of neural networks for improved performance. Practical implications Based on the forecasted energy demand of the DC and an understanding of how the input parameters impact the building's energy demand pattern, neural networks can be deployed to optimize the cooling systems of the DC for reduced energy cost. Originality/value The use of TSA for optimizing the weights and biases of a neural network is a novel study. The application context of this study which is DCs is quite untapped in the literature, leaving many gaps for further research. The proposed prediction model can be further applied to other regression tasks and classification tasks. Another contribution of this study is the analysis of the neural network's input parameters, which provides insight into the level to which each parameter influences the DC’s energy demand profile.

Forecasting the momentum using customised loss function for financial series

2021 ◽  
Vol 14 (4) ◽  
pp. 702-713
Author(s):  
N. Prabakaran ◽  
Rajasekaran Palaniappan ◽  
R. Kannadasan ◽  
Satya Vinay Dudi ◽  
V. Sasidhar
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Loss Function ◽  
Short Term Memory ◽  
Learning Algorithm ◽  
Financial Time Series ◽  
Storage Unit ◽  
Content Type ◽  
Financial Time ◽  
Unit Structure

PurposeWe propose a Machine Learning (ML) approach that will be trained from the available financial data and is able to gain the trends over the data and then uses the acquired knowledge for a more accurate forecasting of financial series. This work will provide a more precise results when weighed up to aged financial series forecasting algorithms. The LSTM Classic will be used to forecast the momentum of the Financial Series Index and also applied to its commodities. The network will be trained and evaluated for accuracy with various sizes of data sets, i.e. weekly historical data of MCX, GOLD, COPPER and the results will be calculated.Design/methodology/approachDesirable LSTM model for script price forecasting from the perspective of minimizing MSE. The approach which we have followed is shown below. (1) Acquire the Dataset. (2) Define your training and testing columns in the dataset. (3) Transform the input value using scalar. (4) Define the custom loss function. (5) Build and Compile the model. (6) Visualise the improvements in results.FindingsFinancial series is one of the very aged techniques where a commerce person would commerce financial scripts, make business and earn some wealth from these companies that vend a part of their business on trading manifesto. Forecasting financial script prices is complex tasks that consider extensive human–computer interaction. Due to the correlated nature of financial series prices, conventional batch processing methods like an artificial neural network, convolutional neural network, cannot be utilised efficiently for financial market analysis. We propose an online learning algorithm that utilises an upgraded of recurrent neural networks called long short-term memory Classic (LSTM). The LSTM Classic is quite different from normal LSTM as it has customised loss function in it. This LSTM Classic avoids long-term dependence on its metrics issues because of its unique internal storage unit structure, and it helps forecast financial time series. Financial Series Index is the combination of various commodities (time series). This makes Financial Index more reliable than the financial time series as it does not show a drastic change in its value even some of its commodities are affected. This work will provide a more precise results when weighed up to aged financial series forecasting algorithms.Originality/valueWe had built the customised loss function model by using LSTM scheme and have experimented on MCX index and as well as on its commodities and improvements in results are calculated for every epoch that we run for the whole rows present in the dataset. For every epoch we can visualise the improvements in loss. One more improvement that can be done to our model that the relationship between price difference and directional loss is specific to other financial scripts. Deep evaluations can be done to identify the best combination of these for a particular stock to obtain better results.

A big data approach to examining social bots on Twitter

2019 ◽  
Vol 33 (4) ◽  
pp. 369-379 ◽  
Author(s):  
Xia Liu
Keyword(s):  
Big Data ◽  
Fixed Effects ◽  
Latent Dirichlet Allocation ◽  
User Generated Content ◽  
Fixed Effects Model ◽  
Data Set ◽  
Information Distortion ◽  
Content Type ◽  
Related Information ◽  
Brand Reputation

Purpose Social bots are prevalent on social media. Malicious bots can severely distort the true voices of customers. This paper aims to examine social bots in the context of big data of user-generated content. In particular, the author investigates the scope of information distortion for 24 brands across seven industries. Furthermore, the author studies the mechanisms that make social bots viral. Last, approaches to detecting and preventing malicious bots are recommended. Design/methodology/approach A Twitter data set of 29 million tweets was collected. Latent Dirichlet allocation and word cloud were used to visualize unstructured big data of textual content. Sentiment analysis was used to automatically classify 29 million tweets. A fixed-effects model was run on the final panel data. Findings The findings demonstrate that social bots significantly distort brand-related information across all industries and among all brands under study. Moreover, Twitter social bots are significantly more effective at spreading word of mouth. In addition, social bots use volumes and emotions as major effective mechanisms to influence and manipulate the spread of information about brands. Finally, the bot detection approaches are effective at identifying bots. Research limitations/implications As brand companies use social networks to monitor brand reputation and engage customers, it is critical for them to distinguish true consumer opinions from fake ones which are artificially created by social bots. Originality/value This is the first big data examination of social bots in the context of brand-related user-generated content.

scholarly journals Multi-Regional Online Car-Hailing Order Quantity Forecasting Based on the Convolutional Neural Network

Information ◽  
2019 ◽  
Vol 10 (6) ◽  
pp. 193 ◽  
Author(s):  
Zihao Huang ◽  
Gang Huang ◽  
Zhijun Chen ◽  
Chaozhong Wu ◽  
Xiaofeng Ma ◽  
...  
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Travel Demand ◽  
Short Term Memory ◽  
Demand Forecasting ◽  
Image Feature ◽  
Support Vector ◽  
Data Set ◽  
Demand Distribution ◽  
Demand Forecasting Model

With the development of online cars, the demand for travel prediction is increasing in order to reduce the information asymmetry between passengers and drivers of online car-hailing. This paper proposes a travel demand forecasting model named OC-CNN based on the convolutional neural network to forecast the travel demand. In order to make full use of the spatial characteristics of the travel demand distribution, this paper meshes the prediction area and creates a travel demand data set of the graphical structure to preserve its spatial properties. Taking advantage of the convolutional neural network in image feature extraction, the historical demand data of the first twenty-five minutes of the entire region are used as a model input to predict the travel demand for the next five minutes. In order to verify the performance of the proposed method, one-month data from online car-hailing of the Chengdu Fourth Ring Road are used. The results show that the model successfully extracts the spatiotemporal features of the data, and the prediction accuracies of the proposed method are superior to those of the representative methods, including the Bayesian Ridge Model, Linear Regression, Support Vector Regression, and Long Short-Term Memory networks.

Patient visit forecasting in an emergency department using a deep neural network approach

Kybernetes ◽  
2019 ◽  
Vol 49 (9) ◽  
pp. 2335-2348 ◽  
Author(s):  
Milad Yousefi ◽  
Moslem Yousefi ◽  
Masood Fathi ◽  
Flavio S. Fogliatto
Keyword(s):  
Neural Network ◽  
Emergency Department ◽  
Linear Regression ◽  
Deep Neural Network ◽  
Short Term Memory ◽  
Demand Forecasting ◽  
Machine Learning Algorithms ◽  
Support Vector ◽  
Neural Network Approach ◽  
Content Type

Purpose This study aims to investigate the factors affecting daily demand in an emergency department (ED) and to provide a forecasting tool in a public hospital for horizons of up to seven days. Design/methodology/approach In this study, first, the important factors to influence the demand in EDs were extracted from literature then the relevant factors to the study are selected. Then, a deep neural network is applied to constructing a reliable predictor. Findings Although many statistical approaches have been proposed for tackling this issue, better forecasts are viable by using the abilities of machine learning algorithms. Results indicate that the proposed approach outperforms statistical alternatives available in the literature such as multiple linear regression, autoregressive integrated moving average, support vector regression, generalized linear models, generalized estimating equations, seasonal ARIMA and combined ARIMA and linear regression. Research limitations/implications The authors applied this study in a single ED to forecast patient visits. Applying the same method in different EDs may give a better understanding of the performance of the model to the authors. The same approach can be applied in any other demand forecasting after some minor modifications. Originality/value To the best of the knowledge, this is the first study to propose the use of long short-term memory for constructing a predictor of the number of patient visits in EDs.

Sign in / Sign up

Export Citation Format

Share Document