Year: 2017
Type: article
Abstract: The aim of this paper is to get some experience with sequence labeling, specifically, assigning tags or labels to each member in the sequences of utterances in conversations from a corpus.Since nowadays predicting single class label or tag is not adequate.Predicting large number of variables that depends on each other is required.In sequence labeling it is often beneficial to optimize the tags assigned to the sequence as a whole rather than treating each tag decision separately.A machine learning technique termed as Conditional Random Fields, which is designed for sequence labeling will be used in order to take advantage of the surrounding context.Conditional random fields (CRFs), is a scheme for building probabilistic models to divide and tag sequence data.With a given a labeled set of data, baseline set of features will be created and the accuracy of the CRF suite model created using those features will be measured.
Authors: Romansha Chopra, Nivedita Singh, Yang Zhenning, N.Ch.S.N. Iyengar
Institution: Vellore Institute of Technology University
Language: English
Topic: Speech and dialogue systems
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W2766012996
