Extraction and Classification of Semantic Relations from News Recommendation

The classification of semantic relation between terms or objects within text is required for a variety of semantic interpreting tasks, such as textual entailment and inquiry answering. In most circumstances, though, attributing a linear semantic relationship between entities/terms is difficult. This work presents a method for categorizing composite semantic relations based on one or more relationships between entities/terms. In contrast to earlier techniques, the proposed model combines a vast commonsense knowledge and understanding of triple connections with machine learning techniques based on lexical and redistributive word embedding properties. To solve the compound semantic relation classification task, we used a distribution of income navigation technique and sequences classification.