30 results about "Churn rate" patented technology

The invention provides a user information display method based on a voice communication platform. The method comprises the following steps of receiving a preset operation command of a common user about an anchor user, wherein the preset operation command is sent by a first terminal corresponding to the common user in the voice communication platform; generating intimacy of the common user about the anchor user according to the operation command; determining an intimacy range where the intimacy belongs; generating user information of the common user corresponding to the intimacy range about the anchor user; and displaying the user information of the common user about the anchor user, reading a personal icon of the anchor user and a personnel icon of the common user and generating an icon corresponding to the intimacy range and including the personnel icon of the anchor user, wherein the user information includes the icon. Through adoption of the method, the churn rate of platform users can be reduced to be helpful for effective management of platform resources. In addition, the invention also provides a user information display system based on the voice communication platform.

The present invention provides a community dynamic identification method, comprising the following steps: obtaining the geographic location information of the base station; establishing a match with the base station according to the geographic location information of the community; establishing a community base station group ID for base stations near and around the community according to the geographic location of the community; Obtain the mobile phone signaling and call data of the active calling users under the community base station group; according to the information of whether the active calling users initiate calls within the time period obtained from the mobile phone signaling information, the loss rate and growth rate are obtained, so as to determine the community belongs to The type of mobility; according to the average call duration and number of calls in the time period obtained from the call data, it can be judged to obtain the community level to which the community belongs. The positive effect of the present invention is: according to the mobile phone signaling data and call data of the community residents, through the optimization algorithm, the manager's community mobility and community hierarchy can be identified in the community, which is convenient for the social manager to carry out targeted management and communication. prevention.

Embodiments presented herein include systems and methods for performing dynamic difficulty adjustments. Furthermore, embodiments disclosed herein perform dynamic difficulty adjustments using a process that may not be perceptible or less perceptible to the user than static and / or existing difficulty adjustment processes. In some embodiments, historical user information is used by a machine learning system to generate a predictive model that predicts expected duration of game play, such as, for example, expected churn rate, retention rate, length of time a user expects to play a game, or relative to previously played games. An indication of the expected game play time of a user of the historical user set. Before or during game play, a predictive model may be applied to information about the user to predict the user's expected duration of game play. Based on the expected duration, the system may then utilize the mapping data repository in some embodiments to determine how to dynamically adjust the game difficulty, such as, for example, changing the value of one or more game play parameters to make portions of the game less difficult.

Provided are a method, a device, terminal equipment and a medium for generating a customer return visit event, applicable to the field of communication technologies. The method comprises: obtaining, according to a buried point code preset in a page, each control triggered by a user in the page (S101); generating a user operation sequence according to a triggering order of each control (S102); obtaining a standard operation sequence that matches the user operation sequence (S103); comparing the user operation sequence with a standard operation sequence to detect whether each control in the standard operation sequence appears in sequence in the user operation sequence (S104); and if each control in the standard operation sequence does not appear in sequence in the user operation sequence, obtaining a login account of the user, and generating a customer return visit event associated with the login account to perform a return visit to the user on the basis of the customer return visit event (S105). By generating a customer return visit event, the scheme realizes a return visit to the user, ensuring that in the follow-up return visits, potential customers can be tracked and retained in time, thereby reducing the churn rate of the customer.

The invention discloses a browse type service awareness indicator prediction method based on multiple label learning. The browse type service awareness indicator prediction method based on multiple label learning solves the problem about how to timely and accurately predict the KQI of the webpage browsing service of a user according to the scene where the user is. The browse type service awareness indicator prediction method based on multiple label learning can make prediction and early warning for good and bad of the service experience in a special scene for the user according to the service awareness historical data of a large number of users, that is, good and bad of the service awareness indicator in different scenes, and is conductive to discovery of the service experience problem as soon as possible so as to make corresponding measures to make improvement, and can effectively reduce the complaint rate and the rate of off-network.

The invention discloses a method and device for determining an access path and a method and system for determining the page churn rate. The method for determining the access path includes the step 1, obtaining records in an accessed log, the step 2, determining a record set provided with the same user identification and determining a target analysis record, the step 3, determining the next target analysis record according to the previous level record obtained by the previous level access website in the current target analysis record, the step 4, using the next target analysis record as the current target analysis record, the step 5, repeating the step 4 until the previous access website in the current target analysis record is recorded as an invalid access website link, and the step 6, constructing a website record. According to the method for determining the access path, due to the fact that the access path contains complete access web pages and web page access sequences, the information amount contained in the access path is much more than the information amount contained in a path pair in the prior art, access behaviors of users can be directly reflected, and consequently the computation cost of website analysis can be effectively lowered by using the access path as the statistics basis.

The present invention provides a user retention probability prediction method, device, prediction server, and storage medium. The method includes: acquiring user loss data of a target service in at least two historical periods; according to the user loss data of the at least two historical periods, Determine the user churn rate of the target business; estimate the user base retention probability of the target business in the future target period under the influence of time according to the user churn rate; obtain the retention probability adjustment parameters of the target user; The retention probability adjustment parameter and the user base retention probability of the target service in the target period are used to predict the retention probability of the target user in the target period. Considering from the perspective of time series, the present invention integrates the user loss rate that will affect the prediction result in time, significantly improves the prediction accuracy of the user retention probability, and improves the prediction effect.