Key technologies of artificial intelligence in electric power customer service

As the demand for customer service continues to increase,more companies are attempting to apply artificial intelligence technology in the field of customer service,enabling intelligent customer service,reducing customer service pressure,and reducing operating costs.Currently,the existing intelligent customer service has a limited degree of intelligence and can only answer simple user questions,and complex user expressions are difficult to understand.To solve the problem of low accuracy of multi-round dialogue semantic understanding,this paper proposes a semantic understanding model based on the fusion of a convolutional neural network(CNN)and attention.The model builds an“intention-slot”joint model based on the“encoding–decoding”framework and uses hidden semantic information that combines intent recognition and slot filling,avoiding the problem of information loss in traditional isolated tasks,and achieving end-to-end semantic understanding.Additionally,an improved attention mechanism based on CNNs is introduced in the decoding process to reduce the interference of redundant information in the original text,thereby increasing the accuracy of semantic understanding.Finally,by applying the model to electric power intelligent customer service,we verified through an experimental comparison that the proposed fusion model improves the performance of intent recognition and slot filling and can improve the user experience of electric power intelligent customer services.

Measuring Similarity of Academic Articles with Semantic Profile and Joint Word Embedding

Long-document semantic measurement has great significance in many applications such as semantic searchs, plagiarism detection, and automatic technical surveys. However, research efforts have mainly focused on the semantic similarity of short texts. Document-level semantic measurement remains an open issue due to problems such as the omission of background knowledge and topic transition. In this paper, we propose a novel semantic matching method for long documents in the academic domain. To accurately represent the general meaning of an academic article, we construct a semantic profile in which key semantic elements such as the research purpose, methodology, and domain are included and enriched. As such, we can obtain the overall semantic similarity of two papers by computing the distance between their profiles. The distances between the concepts of two different semantic profiles are measured by word vectors. To improve the semantic representation quality of word vectors, we propose a joint word-embedding model for incorporating a domain-specific semantic relation constraint into the traditional context constraint. Our experimental results demonstrate that, in the measurement of document semantic similarity, our approach achieves substantial improvement over state-of-the-art methods, and our joint word-embedding model produces significantly better word representations than traditional word-embedding models.