Text knowledge representation model based on human concept learning

The essential abilities of text knowledge representation, such as automatic construction, carrying abundant semantics and flexible reasoning, should be held due to the rapid growth of web resources and the requirements of the reasoning-based web services. However, current text knowledge representation models either lose many textual semantics or cannot be constructed automatically. To solve the above issues, text knowledge representation model is proposed based on the concept algebra of human concept learning. Then, the degree-2 power series hypothesis is developed and the reasoning ability of text representation is proposed. Finally, the results compared with current knowledge representation models show that our model performs better than other models in representing text knowledge.