Enhancing rating prediction for recommendation by metric learning with gray relational analysis

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Jiangmei Chen ◽  
Wende Zhang ◽  
Qishan Zhang
Keyword(s):  
Metric Learning ◽  
Prediction Algorithm ◽  
Gray Relational Analysis ◽  
Rating Data ◽  
Data Sparsity ◽  
Content Type ◽  
Relational Analysis ◽  
Rating Prediction ◽  
Real World Datasets ◽  
The Impact

PurposeThe purpose of the paper is to improve the rating prediction accuracy in recommender systems (RSs) by metric learning (ML) method. The similarity metric of user and item is calculated with gray relational analysis.Design/methodology/approachFirst, the potential features of users and items are captured by exploiting ML, such that the rating prediction can be performed. In metric space, the user and item positions can be learned by training their embedding vectors. Second, instead of the traditional distance measurements, the gray relational analysis is employed in the evaluation of the position similarity between user and item, because the latter can reduce the impact of data sparsity and further explore the rating data correlation. On the basis of the above improvements, a new rating prediction algorithm is proposed. Experiments are implemented to validate the effectiveness of the algorithm.FindingsThe novel algorithm is evaluated by the extensive experiments on two real-world datasets. Experimental results demonstrate that the proposed model achieves remarkable performance on the rating prediction task.Practical implicationsThe rating prediction algorithm is adopted to predict the users' preference, and then, it provides personalized recommendations for users. In fact, this method can expand to the field of classification and provide potentials for this domain.Originality/valueThe algorithm can uncover the finer grained preference by ML. Furthermore, the similarity can be measured using gray relational analysis, which can mitigate the limitation of data sparsity.

The importance and the adaptation of internal resources as a competitive advantage for the internationalization of food companies

2019 ◽  
Vol 9 (3) ◽  
pp. 305-320
Author(s):  
Gessuir Pigatto ◽  
Giuliana Aparecida Santini Pigatto ◽  
Eduardo Guilherme Satolo ◽  
Amanda dos Santos Negreti
Keyword(s):  
Competitive Advantage ◽  
Design Methodology ◽  
Gray Relational Analysis ◽  
Resource Based View ◽  
Content Type ◽  
Internal Resources ◽  
Relational Analysis ◽  
Gray Correlation ◽  
Correlation System ◽  
Practical Implications

Purpose The purpose of this paper is to analyze how Brazilian food companies in the State of São Paulo determine the importance of and the need to adapt their internal resources as a competitive advantage for internationalization. Design/methodology/approach From a resource-based view (RBV), 35 different factors grouped into four categories were identified and presented to 24 companies. The data were analyzed through a gray relational analysis to establish all factors’ order of importance. Findings Factors linked to human and organizational resources present greater adaptability and allow companies competitive and sustainable advantages but have not yet been explored thoroughly. Identifying and adapting internal resources do not guarantee achieving competitive and sustainable advantages, as the access to international market is also a consequence of commercial agreements developed by countries and economic blocks. Practical implications The analysis highlights the fragility of competitiveness among the companies analyzed in exporting products with commodity characteristics, with none or little differentiation. Such products are traded mainly through trading companies, which allow the access of the same market to internal competitors and other countries. Thus, any lapse promoted by the company may be enough for it to lose its competitiveness and, hence, market space. Originality/value This paper stands out in the field of strategic management, specifically in the research on RBV, exportation and competitiveness, by making use of the theory of gray correlation system in an innovative and original way.

The Impact Factors Analysis of the Continuous Safety of the Yangtze River Shipping

2013 ◽  
Vol 655-657 ◽  
pp. 2279-2283
Author(s):  
Lu Wang ◽  
Qing Liu ◽  
Kai Jin Xu ◽  
Xiao Li Xu
Keyword(s):  
Continuous Improvement ◽  
Yangtze River ◽  
Safety Management ◽  
Gray Relational Analysis ◽  
Impact Factors ◽  
The Yangtze River ◽  
Safety Standards ◽  
Relational Analysis ◽  
Factors Analysis ◽  
The Impact

This paper aims to analyze factors affect the continuous safety of the Yangtze River shipping, it studied the connotation and feature of continuous safety, and based on which selecting safety input, human behavior, safety management, safety state, safety culture as the main impact aspects, using the gray relational analysis to extract the key impact factors belong to the abovementioned five aspects. The result shows that safety management, safety input and safety state have bigger impact on sustainable safety state and continuous improvement of system safety standards.

A Collaborative Filtering Recommendation Algorithm Based on Cloud Model, Item Attribute and User Data

2014 ◽  
Vol 610 ◽  
pp. 747-751
Author(s):  
Jian Sun ◽  
Xiao Ying Chen
Keyword(s):  
Collaborative Filtering ◽  
Prediction Accuracy ◽  
Cloud Model ◽  
Prediction Algorithm ◽  
Recommendation Algorithm ◽  
Rating Data ◽  
User Data ◽  
Item Attribute ◽  
Rating Prediction ◽  
User Ratings

Aiming at the problems of extremely sparse of user-item rating data and poor recommendation quality, we put forward a collaborative filtering recommendation algorithm based on cloud model, item attribute and user data which combined with the existing literatures. A rating prediction algorithm based on cloud model and item attribute is proposed, based on idea that the similar users rating for the same item are similar and the same user ratings for the similar items are similar and stable. Through compare and analysis this paper’s and other studies experimental results, we get the conclusion that the rating prediction accuracy is improved.

scholarly journals Multi-objective optimization of carbon/glass hybrid composites with newly developed resin (NDR) using gray relational analysis

2020 ◽  
Vol 16 (6) ◽  
pp. 1709-1729
Author(s):  
Sagar Dnyandev Patil ◽  
Yogesh J. Bhalerao
Keyword(s):  
Mechanical Properties ◽  
Composite Laminates ◽  
Hybrid Composites ◽  
Research Work ◽  
Optimum Combination ◽  
Gray Relational Analysis ◽  
Stacking Sequence ◽  
Content Type ◽  
Relational Analysis ◽  
Design Variables

PurposeIt is seen that little amount of work on optimization of mechanical properties taking into consideration the combined effect of design variables such as stacking angle, stacking sequence, different resins and thickness of composite laminates has been carried out. The focus of this research work is on the optimization of the design variables like stacking angle, stacking sequence, different resins and thickness of composite laminates which affect the mechanical properties of hybrid composites. For this purpose, the Taguchi technique and the method of gray relational analysis (GRA) are used to identify the optimum combination of design variables. In this case, the effect of the abovementioned design variables, particularly of the newly developed resin (NDR) on mechanical properties of hybrid composites has been investigated.Design/methodology/approachThe Taguchi method is used for design of experiments and with gray relational grade (GRG) approach, the optimization is done.FindingsFrom the experimental analysis and optimization study, it was seen that the NDR gives excellent bonding strength of fibers resulting in enhanced mechanical properties of hybrid composite laminates. With the GRA method, the initial setting (A3B2C4D2) was having GRG 0.866. It was increased by using a new optimum combination (A2B2C4D1) to 0.878. It means that there is an increment in the grade by 1.366%. Therefore, using the GRA approach of analysis, design variables have been successfully optimized to achieve enhanced mechanical properties of hybrid composite laminates.Originality/valueThis is an original research work.

scholarly journals Watershed Drought and Ecosystem Services: Spatiotemporal Characteristics and Gray Relational Analysis

2021 ◽  
Vol 10 (2) ◽  
pp. 43
Author(s):  
Jizhou Bai ◽  
Zixiang Zhou ◽  
Yufeng Zou ◽  
Bakhtiyor Pulatov ◽  
Kadambot H. M. Siddique
Keyword(s):  
Ecosystem Services ◽  
Soil Erosion ◽  
Vegetation Cover ◽  
Soil Conservation ◽  
Swat Model ◽  
Vegetation Restoration ◽  
Gray Relational Analysis ◽  
Relational Analysis ◽  
Spatiotemporal Characteristics ◽  
The Impact

This study explored the spatiotemporal characteristics of drought and ecosystem services (using soil conservation services as an example) in the YanHe Watershed, which is a typical water basin in the Loess Plateau of China, experiencing soil erosion. Herein, soil conservation was simulated using the Soil and Water Assessment Tool (SWAT), and the relationship between drought, soil conservation services, and meteorological, vegetation, and other factors since the implementation of the ‘Grain for Green’ Project (GFGP) in 1999, were analyzed using the gray relational analysis (GRA) method. The results showed that: (1) The vegetation cover of the Watershed has increased significantly, and evapotranspiration (ET) increased by 14.35 mm·a−1, thereby increasing water consumption by 8.997 × 108 m3·a−1 (compared to 2000). (2) Drought affected 63.86% of the watershed area, gradually worsening from south to north; it decreased in certain middle areas but increased in the humid areas on the southern edge. (3) The watershed soil conservation services, measured by the soil conservation modulus (SCM), increased steadily from 116.87 t·ha−1·a−1 in 2000 to 412.58 t·ha−1·a−1 in 2015, at a multi-year average of 235.69 t·ha−1·a−1, and indicated great spatial variations, with a large variation in the downstream and small variations in the upstream and midstream areas. (4) Integrating normalized difference vegetation index (NDVI) data into SWAT model improved the model simulation accuracy; during the calibration period, the coefficient of determination (R2) increased from 0.63 to 0.76 and Nash–Sutcliffe efficiency (NSE) from 0.46 to 0.51; and during the validation period, the R2 increased from 0.82 to 0.93 and the NSE from 0.57 to 0.61. (5) The GRA can be applied to gray control systems, such as the ecosystem; herein, vegetation cover and drought primarily affected ET and soil conservation services. The analysis results showed that vegetation restoration enhanced the soil conservation services, but increased ET and aggravated drought to a certain extent. This study analyzed the spatiotemporal variations in vegetation coverage and the response of ET to vegetation restoration in the YanHe Watershed, to verify the significant role of vegetation restoration in restraining soil erosion and evaluate the extent of water resource consumption due to ET in the semi-arid and semi-humid Loess-area basin during the GFGP period. Thus, this approach may effectively provide a scientific basis for evaluating the ecological effects of the GFGP and formulating policies to identify the impact of human ecological restoration on ecosystem services.

scholarly journals Metric Factorization with Item Cooccurrence for Recommendation

Symmetry ◽  
2020 ◽  
Vol 12 (4) ◽  
pp. 512 ◽  
Author(s):  
Honglin Dai ◽  
Liejun Wang ◽  
Jiwei Qin
Keyword(s):  
Matrix Factorization ◽  
Euclidean Distance ◽  
Metric Learning ◽  
Sparse Data ◽  
Inner Product ◽  
Interaction Matrix ◽  
Latent Factors ◽  
Rating Prediction ◽  
Real World Datasets ◽  
Item Item

In modern recommender systems, matrix factorization has been widely used to decompose the user–item matrix into user and item latent factors. However, the inner product in matrix factorization does not satisfy the triangle inequality, and the problem of sparse data is also encountered. In this paper, we propose a novel recommendation model, namely, metric factorization with item cooccurrence for recommendation (MFIC), which uses the Euclidean distance to jointly decompose the user–item interaction matrix and the item–item cooccurrence with shared latent factors. The item cooccurrence matrix is obtained from the colike matrix through the calculation of pointwise mutual information. The main contributions of this paper are as follows: (1) The MFIC model is not only suitable for rating prediction and item ranking, but can also well overcome the problem of sparse data. (2) This model incorporates the item cooccurrence matrix into metric learning so it can better learn the spatial positions of users and items. (3) Extensive experiments on a number of real-world datasets show that the proposed method substantially outperforms the compared algorithm in both rating prediction and item ranking.

Adversarial auto-encoder for rating prediction with ratings and reviews

2021 ◽  
pp. 1-10
Author(s):  
Jin Yi ◽  
Jiajin Huang ◽  
Jin Qin
Keyword(s):  
Real World ◽  
Empirical Studies ◽  
Auxiliary Information ◽  
Rating Data ◽  
Adversarial Learning ◽  
Rich Information ◽  
Rating Prediction ◽  
Sparsity Problem ◽  
Real World Datasets

Recommender systems have been widely used in our life in recent years to facilitate our life. And it is very important and meaningful to improve recommendation performance. Generally, recommendation methods use users’ historical ratings on items to predict ratings on their unrated items to make recommendations. However, with the increase of the number of users and items, the degree of data sparsity increases, and the quality of recommendations decreases sharply. In order to solve the sparsity problem, other auxiliary information is combined to mine users’ preferences for higher recommendation quality. Similar to rating data, review data also contain rich information about users’ preferences on items. This paper proposes a novel recommendation model, which harnesses an adversarial learning among auto-encoders to improve recommendation quality by minimizing the gap of the rating and review relation between a user and an item. The empirical studies on real-world datasets show that the proposed method improves the recommendation performance.

Determinants of banking sector’s credit granting policy for the yacht industry in Taiwan

2016 ◽  
Vol 1 (1) ◽  
pp. 55-75 ◽  
Author(s):  
Yi-Chih Yang ◽  
Hsien-Pin Liu
Keyword(s):  
Design Methodology ◽  
Banking Sector ◽  
Bank Credit ◽  
Gray Relational Analysis ◽  
Content Type ◽  
Relational Analysis ◽  
Priority Ranking ◽  
Research Findings ◽  
Ranking Order ◽  
Credit Granting

Purpose This paper aims to investigate bank credit policies and uncover yacht building finance assessment factors from bank credit policies toward the yacht industry. Design/methodology/approach This study’s questionnaire attempts to identify survey respondents’ degrees of awareness through difference analysis, and then uses entropy weighting and gray relational analysis to discover priority ranking order of bank credit assessment considerations from the perspective of Taiwan’s banking sector. Findings The research findings show that yacht builders have to review their ship financing application methods and improve shortcomings to meet banks’ credit granting requirements. Originality/value Banks emphasize yacht builders’ repayment ability to protect their depositors and shareholders.

Predicting corporate credit rating based on qualitative information of MD&A transformed using document vectorization techniques

2020 ◽  
Vol 54 (2) ◽  
pp. 151-168
Author(s):  
Jinwook Choi ◽  
Yongmoo Suh ◽  
Namchul Jung
Keyword(s):  
Credit Rating ◽  
Prediction Models ◽  
Quantitative Information ◽  
Qualitative Information ◽  
Rating Data ◽  
Content Type ◽  
Financial Variables ◽  
Corporate Credit ◽  
Rating Prediction ◽  
Corporate Credit Rating

PurposeThe purpose of this study is to investigate the effectiveness of qualitative information extracted from firm’s annual report in predicting corporate credit rating. Qualitative information represented by published reports or management interview has been known as an important source in addition to quantitative information represented by financial values in assigning corporate credit rating in practice. Nevertheless, prior studies have room for further research in that they rarely employed qualitative information in developing prediction model of corporate credit rating.Design/methodology/approachThis study adopted three document vectorization methods, Bag-Of-Words (BOW), Word to Vector (Word2Vec) and Document to Vector (Doc2Vec), to transform an unstructured textual data into a numeric vector, so that Machine Learning (ML) algorithms accept it as an input. For the experiments, we used the corpus of Management’s Discussion and Analysis (MD&A) section in 10-K financial reports as well as financial variables and corporate credit rating data.FindingsExperimental results from a series of multi-class classification experiments show the predictive models trained by both financial variables and vectors extracted from MD&A data outperform the benchmark models trained only by traditional financial variables.Originality/valueThis study proposed a new approach for corporate credit rating prediction by using qualitative information extracted from MD&A documents as an input to ML-based prediction models. Also, this research adopted and compared three textual vectorization methods in the domain of corporate credit rating prediction and showed that BOW mostly outperformed Word2Vec and Doc2Vec.

Sign in / Sign up

Export Citation Format

Share Document