| CPC G06Q 50/205 (2013.01) | 5 Claims |
|
1. A method for managing school attendance of a smart city based on Internet of Things, which is implemented by a processor of a school attendance management platform, the method comprising:
obtaining student registration information based on a student user platform, the student registration information including an intention for school attendance of a student;
obtaining a student evaluation score based on a plurality of school user sub-platforms;
aggregating the student registration information and the student evaluation score through a school attendance service platform;
for each student, determining a ranking score of the student at an interest school based on a weighted sum of a distance score, the student evaluation score, and a draw score of the student, wherein weights of the distance score, the student evaluation score, and the draw score are determined based on an estimated time of arrival (ETA) between a house of the student and the interest school and a performance of the student, the draw score is determined based on a random number selected by the student, the random number and the corresponding draw score are randomly generated and stored in a blockchain, the blockchain stores a decryption key; when the random number is selected by the student, sending the decryption key to the student user platform, such that based on the decryption key of the school attendance management platform, the student obtains the draw score corresponding to the random number by the student user platform, and determines, by the student user platform, whether the decryption key of the school attendance management platform is tempered by comparing the decryption key obtained from the school attendance management platform with the decryption key obtained from the blockchain to verify the reliability of the draw score;
determining a ranking of the student at the interest school based on the ranking score, including:
determining, separately, a student selection probability of each higher-ranked student based on a machine learning model, and obtaining a probability that a lower-ranked student moves forward, wherein the machine learning model is a logistic regression model; and the machine learning model is obtained through a training process by the processor of the school attendance management platform, the training process comprising:
receiving a historical data from a storage device, wherein the historical data is a historical school place allocation data obtained from one or more of the school user sub-platforms, the student user platform, the school attendance management platform, and the school attendance service platform, and the historical school place allocation data includes historical intention for school attendance of the student, historical distance scores, historical ranking of the interest school, historical student evaluation scores;
generating a labeled historical data by labeling a corresponding student selection result (yes or no) in the historical data;
inputting the labeled historical data into an initial machine learning model for training to obtain a trained machine learning model;
predicting the student selection probability by processing the intention for school attendance of the student, the distance score, the ranking of the student at the interest school, and the student evaluation score using the trained machine learning model, determining a ranking advance probability based on the student selection probability, and publishing the ranking advance probability to the student;
obtaining a selection result made by the student for ranking candidate schools based on the student user platform and the ranking advance probability; wherein the ranking candidate schools include at least one school with the highest ranking score corresponding to the student; and the selection result includes at most selecting one of the ranking candidate schools;
aggregating selection results of students based on the school attendance service platform and sending the selection results to the school attendance management platform;
determining school place allocation in conjunction with the selection results;
based on the school place allocation, obtaining distance information between houses of the students and schools from a geographic information platform, and determining whether the school place allocation meets an evaluation index in combination with the performance of the students; and
adjusting a parameter of the school place allocation in response to the school place allocation not meeting the evaluation index.
|