Sequence Labeling Using Deep Neural Nets
Nikhila N.V. · International Journal of Advanced Trends in Computer Science and Engineering · 2019
Now a day's sequence labeling has become most interesting topic in the current technical era.Sequence labeling is a type of pattern recognition task that involves the algorithmic assignment of a categorical label to each member of sequence of observed values, and also, it is treated as an independent task.By using traditional methods such as HMM and CRF we can implement sequence labeling.Both the methods take the sequence of input and learn to predict an optimal sequence of labels.These are very powerful methods, but they have not experienced the great success due to some drawbacks like lack of semantic awareness and can't handle longer sequential dependencies.So by using deep learning techniques such as recurrent neural networks, they can capture the local dependencies and find longer patterns.Real world applications where the sequence labeling can be applied are Google search Engine.Where in the search box if we type some words automatically Google will suggest some sentences or words which makes our work easier.