A cognitive service-oriented user intent recognition method and system

Abstract

The invention discloses a cognitive service-oriented user intention recognition method and system. The method includes the following steps: step S1, collecting and labeling a large-scale corpus; step S2, mining concept co-occurrence patterns; step S3, constructing concept co-occurrence patterns The semantic model is now enhanced; step S4, implementing an intent recognition scheme. The present invention proposes the concept of "concept co-occurrence pattern" in medical query text, excavates frequent co-occurrence patterns from medical query records and uses this co-occurrence pattern to identify user intentions, and reveals the relationship between concept co-occurrence and user intentions. potential associations and reveal their significant impact on the problem of multi-intent and implicit intent recognition. The present invention fully considers the complexity and ambiguity of the user query text expression, extracts features from two dimensions of semantic features and concept co-occurrence features for intent recognition, solves complex intent recognition problems such as multiple intents and implicit intents, and solves complex intent recognition problems such as multiple intents and implicit intents. It has good transferability on the problem of intent recognition.

Description

technical field [0001] The invention belongs to the technical field of computer services, and relates to a cognitive service-oriented user intent recognition method and system, in particular to an intent recognition method and system based on a conceptual co-occurrence enhanced semantic model. Background technique [0002] Cognitive services are a bridge between humans and machines, and are widely used in question answering systems, human-machine dialogue systems, and smart terminals. User intent recognition has been extensively studied in recent years as an important module of cognitive services. But most studies are conducted in simple life scenarios. Usually, user statements are simple and straightforward. Different from conventional application scenarios, there are many implicit intent and multi-intent scenarios in the medical query field, and words of different intent categories often co-occur in medical-related query sentences. Traditional user intent recognition mo...

Images