Method for identifying employment place by using k-means clustering algorithm

Abstract

The invention discloses a method for identifying an employment place by using a k-means clustering algorithm, and belongs to the field of urban employment place research. According to the method, a Thiessen polygon formed by a mobile phone base station is used as a to-be-identified unit; data statistics is performed on the occurrence time points of the access activities of any to-be-identified unit in the two working days; and the user behavior is judged by comparing the base station position information recorded by each piece of mobile phone signaling data of the user with the base station position information recorded before and after. The latest leaving time node, the earliest entering time node and the staying time length of all users of each to-be-identified unit are counted, the centroid distribution rule of each to-be-identified unit is analyzed by using a K-means clustering method, and the centroid distribution rule of each to-be-identified unit is calculated according to the centroid number meeting the staying time length threshold value and the positive and negative directions of the staying time length, and whether the to-be-identified unit is an employment place and the type of the to-be-identified unit is judged according to the overlapping / interval of the access time of each centroid.

Description

Technical field [0001] The present invention relates to the field of urban employment, and in particular, a method of identifying employment using a K mean clustering algorithm. Background technique [0002] Employment is a fixed spatial range composed of urban residents to participate in production activities. Time, in terms of employment participation in production activities, usually present more people, there are fewer people in the evening; location fixability is residents Frequent space for production activities. Therefore, according to the above characteristics, the employment of this study refers to the location of urban residents involved in the production activities. [0003] Traditional employment recognition ideas mainly depends on the regularity of users during the daytime, that is, the number of times the number of times the different time nodes in one cycle will be judged by the number of different time nodes during the day, and it is easy to employ, there is certa...

AI Technical Summary

Problems solved by technology

[0004] At present, the biggest problem with the above-mentioned identification method is: first, the identification time mainly depends on the mobile base station of the user in a single period (daytime), and it is likely to be night shift, double shift and third shift (such as hospitals and factories and other specific enterprises) Institutions) ignore the actual place of employment, and even misjudge it as the user’s place of residence; secondly, this method directly judges the place of employment from the perspective of the user’s individual behavior data. The user's trips other than employment are limited, and there is a lack of statistical analysis of a large number of user data in the micro-spatial unit to accurately determine whether the micro-spatial unit is a place of employment.

Images