DEEP LEARNING MODEL FOR SIMILAR QUESTION IDENTIFICATION ON QUORA QUESTIONS USING LSTM
Abstract
A single question can be structured in many different ways, and answering the same question repeatedly is not a wise solution. Filtering these types of questions manually will be monotonous and mindless. Quora is a question-answer platform that has traffic of millions of users, and this increases the probability of questions with similar semantic meaning being asked repeatedly. In this paper we compare different types of machine learning and deep learning algorithms to identify the best algorithm. Log loss metric will be used to measure the performance of the algorithms. The algorithms are tested on the Quora question dataset which is available on Kaggle.
Keywords
Quora, Semantic meaning, Log loss metric, Kaggle, Machine Learning, Deep Learning