2017/07/04 by Zsolt Bikádi, Zsolt Bikadi, Sapumal Ahangama +4
Computer Science · Mathematics · #Artificial intelligence #Biology #Computational biology #Computer science #Computers and Society (cs.CY) #Data mining #Domain (mathematical analysis) #Domain name #FOS: Computer and information sciences #Machine learning #Mathematical analysis #Mathematics #Pattern recognition (psychology) #Support vector machine #Telecommunications #Text and Document Classification Technologies #Throughput #Topic Modeling #Web Data Mining and Analysis #World Wide Web #cs.CY
paper · pdf · doi:10.48550/arxiv.1707.00906
arxiv created 2017/07/04 · openalex publication_date 2017/07/04 · arxiv updated 2017/07/05 · openalex created_date 2017/07/14 · openalex updated_date 2026/07/28
As connected devices multiply and the internet matures into a ubiquitous platform for exchange and communication, the question of what makes a domain name valuable is ever more significant. Due to the scarcity of meaningful vocabulary and the persistence of domain-related data, the buying and selling of previously owned domain names, also known as the domain aftermarket, has evolved into a billion dollar industry. Each day over a 100,000 domain names expire and become available for re-registration. Manual appraisal is impossible at such a volume; thus a method for the automated identification of valuable domain names is called for. The aim of our study was to develop a method for high throughput screening of domain names for rapid identification of the valuable ones. Five different aspects that make a domain name valuable were identified: name quality, domain authority, domain traffic, active domain age and domain health. An SVM method was developed for high throughput screening of domain names. Our method was able to identify valuable domain names with 97% accuracy for the test set and 93% for the external set and can be used for routinely screening the domain aftermarket.