Multi label restaurant classification using support vector machine
| Autor: | Azhar F. Al-zubidi, Ehsan Qahtan, Aseel B. Alnajjar, Nadia F. AL-Bakri |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Multi-label classification
General Computer Science business.industry Computer science Mechanical Engineering 0211 other engineering and technologies Biomedical Engineering Bioengineering 02 engineering and technology Machine learning computer.software_genre Industrial and Manufacturing Engineering Set (abstract data type) Support vector machine ComputingMethodologies_PATTERNRECOGNITION Architecture 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing The Internet Artificial intelligence Electrical and Electronic Engineering business computer 021101 geological & geomatics engineering |
| Zdroj: | Periodicals of Engineering and Natural Sciences (PEN). 9:774 |
| ISSN: | 2303-4521 |
| DOI: | 10.21533/pen.v9i2.1876 |
| Popis: | Many internet websites are hosted with a vast amount of information about restaurants which are not identified properly according to some predefined features to fit users’ interests. Thus, restaurant classification was needed to solve this problem. Restaurant classification has become very important for individuals and food business applications to spread their services via the Internet. In this paper, a modest model is proposed to classify restaurants based on their predefined features which are used as factors affecting restaurant's ratings. The usage of multi label classification is utilized for labelling to maintain acceptable requirements for restaurant's services. Two proposed labels are suggested resulted from the output of two classifiers each operate on a specific set of features. Support vector machine is used for classification because of its effectiveness in restaurant's label separation. The final prediction label is yielded after applying the proposed hypothesis rules. The experimental results conducting Zomato dataset show that the proposed multi label model achieved approximately about 88% for prediction accuracy. Using the proposed model for classification had led to get a collection of accepted restaurants according to user favorites. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|