HASKER: An efficient algorithm for string kernels. Application to polarity classification in various languages

• Published in: Procedia computer science Vol. 112; pp. 1755 - 1763
• Main Authors: Popescu, Marius, Grozea, Cristian, Tudor Ionescu, Radu
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 2017
• Subjects: blended specturm kernel; intersection kernel; kernel methods; open-source code; opining mining; polarity classification; sentiment analysis; similarity-based learning; string kernels; string kernels tool; string kernels; similarity-based learning; opining mining; sentiment analysis; open-source code; intersection kernel; string kernels tool; kernel methods; polarity classification; blended specturm kernel
• ISSN: 1877-0509; 1877-0509
• DOI: 10.1016/j.procs.2017.08.207

String kernels have successfully been used for various NLP tasks, ranging from text categorization by topic to native language identification. In this paper, we present a simple and efficient algorithm for computing various spectrum string kernels. When comparing two strings, we store the p-grams in the first string into a hash table, and then we apply a hash table lookup for the p-grams that occur in the second string. In terms of time, we show that our algorithm can outperform a state-of-the-art tool for computing string similarity. In terms of accuracy, we show that our approach can reach state-of-the-art performance for polarity classification in various languages. Our efficient implementation is provided online for free at http://string-kernels.herokuapp.com.