Abstract: Objective To establish an online consultation platform for disease prevention and control institutions,analyze the consulting data to understand the public health needs,and provide references for optimizing the allocation of public health resources. Methods Construct an intelligent dialogue robot and apply it to the WeChat public account of Nanjing Disease Control and Prevention to provide public health information consultation services.Apply the joinpoint regression(JPR) model to calculate the daily percent change(DPC) and the average daily percent change(ADPC),assessing the trend of the robot's response capability over time.Analyze the distribution of consultation information by time,disease type,and topic,and present it visually. Results From June 2023 to August 2024,users submitted a total of 25 766 consultations,averaging 58 per day.These were categorized and merged to form a list of 294 distinct requests.The robot's daily effective response rate was 97.9%(96.1%-100%),showing a trend of rapid increase in the early stage(DPC=2.43%,P<0.01) and a high-level steady state in the later period(DPC=0,P<0.01).The robot's daily precise response rate was 64.0%(60.0%-69.0%),exhibiting a trend of rapid increase in the early stage(DPC=6.58%,P<0.01),slow increase in the middle stage(DPC=0.14%,P<0.01),and a high-level steady state in the later period(DPC=0,P>0.05).The volume of inquiries varied across different weekdays and time slots,with the highest concentration of consultations occurring between 8 to 16 o'clock on Mondays and 10 to 14 o'clock on Tuesdays.The most frequently consulted diseases included AIDS(3 334,28.5%),rabies(2 599,22.2%),HPV infection(2 369,20.3%),hepatitis B(933,8.0%),chickenpox and herpes zoster(267,2.3%),collectively accounting for 82.6% of the total consultation volume.The top five categories of inquiries were vaccination(7 900,52.3%),testing and consultation(2 701,17.9%),health check-ups(1 749,11.6%),health education(1 389,9.2%),and disease prevention and control measures(550,3.6%),together accounting for 94.6% of the total. Conclusion Intelligent online consultation has become an effective way to answer public health inquiries and understand public health needs.Priority should be given to key diseases and key time periods in the allocation of public health resources. Key areas where public health resources should be allocated more include immunization planning projects,infectious disease screening,health education,public health service informatization,and the improvement of cross-departmental collaborative working mechanism.

Key words: Intelligent dialogue robot, Online consultation, Public health services, Demand analysis, Resource allocation