Named Entity Recognition for English Tweets using Random Kitchen Sink Algorithm

The information obtained electronically is vast and difficult for users to access the exact information within permissible time. The various Natural Language Processing (NLP) task has been carried out in the �eld of Arti�cial Intelligence (AI) to extract structured information from this large amount of unstructured data. Named Entity Recognition (NER) is one such Information Extraction (IE) system which identi�es the elements such as name of person, location, organization, quantities, time expressions etc. and classify into set of pre-de�ned classes. In this thesis, language independent system is developed using Machine Learning algorithms such as Conditional Random Field (CRF), Support Vector Machine (SVM) and Random Kitchen Sink (RKS). A unique approach has been carried out in implementing Random Kitchen Sink algorithm for Named Entity Recognition. This paper deals with RKS for English tweets. The accuracy for this system is obtained as 82.60%.