CN107333285A - A kind of method that cellular signal strength is predicted according to surfing Internet with cell phone daily record - Google Patents
A kind of method that cellular signal strength is predicted according to surfing Internet with cell phone daily record Download PDFInfo
- Publication number
- CN107333285A CN107333285A CN201710544918.6A CN201710544918A CN107333285A CN 107333285 A CN107333285 A CN 107333285A CN 201710544918 A CN201710544918 A CN 201710544918A CN 107333285 A CN107333285 A CN 107333285A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- imsi
- httplog
- triple
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/06—Testing, supervising or monitoring using simulated traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/10—Scheduling measurement reports ; Arrangements for measurement reports
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides the method that one kind predicts cellular signal strength according to surfing Internet with cell phone daily record (httplog data), its step is:IMSI and time field in httplog data, corresponding user's triple is found in MME data, can be in the corresponding signal strength data of MR extracting datas by user's triple, signal strength data and httplog data are combined into training data, carry out data statistics, feature extraction, model construction, model training, finally it is predicted using the model trained, according to httplog data prediction cellular signal strengths.
Description
Technical field
The invention belongs to areas of information technology.It is specifically related to a kind of that cellular signal strength is predicted according to surfing Internet with cell phone daily record
Method.
Background technology
With the development of mobile Internet, increasing people's selection uses surfing Internet with cell phone, for mobile operator,
By the statistics and analysis to a large amount of internet logs, the price bidding of user is grasped, the access behavior hobby of user can be speculated,
And then carry out accurately advertisement putting and marketing.User behavior and base station case can also be combined, the network optimization is carried out, improves and uses
Experience at family.During the network optimization, the signal intensity of user mobile phone is particularly important data, but it is only stored in MR data
In.MR data refer to that information sends a data per 480ms (470ms on signaling channel) on a traffic channel, that is to say, that appoint
In the mobile phone of open state, often 480ms sends a data to base station for where, therefore can produce a large amount of MR data daily, and it is stored
Cost is very big, and operator is difficult that MR data are all stored carry out statistics and analysis.Prior art can only pass through MR data
Signal intensity is obtained, no other technologies predict hand to determine cellular signal strength by surfing Internet with cell phone daily record data
Machine signal intensity is particularly important, and can both save carrying cost, again can be combined MR data with internet log data
Carry out the network optimization.
The content of the invention
It is an object of the invention to provide a kind of cellular signal strength Forecasting Methodology based on surfing Internet with cell phone daily record data, pass through
This method predicts the signal intensity of active user's mobile phone.
The method that cellular signal strength is predicted according to surfing Internet with cell phone daily record that the present invention is provided, detailed step includes:
1st, the httplog data of all mobile phones of current period are obtained, IMSI and time field is extracted;
Define 1:Httplog data;
When httplog data refer to that user uses mobile network, the daily record data that base station is preserved.What the data were included
Information has:
(1) user profile:Mainly include the essential information of user mobile phone, as phone number, mobile phone string number, used in it is clear
Look at device, data traffic size etc. for downloading and uploading.
(2) base station information:The main essential information that base station is connected including user, as where latitude and longitude of base station, base station
Administrative region, base station IDs etc..
(3) behavioural information:Refer to the specific internet behavior of user, the network address such as accessed, the application type used, http transmission
Field type etc..
(4) temporal information:When referring to user's generation internet behavior, the information relevant with the time, such as:At the beginning of internet behavior
Between and end time, TCP link setup response times, service delay etc..
Define 2:IMSI fields;
IMSI fields refer to international mobile subscriber identity (International Mobile Subscriber
Identification Number), it is the mark for distinguishing mobile subscriber, is stored in SIM card, is to be used to distinguish mobile subscriber
Effective information.IMSI fields total length be no more than 15, equally using 0~9 numeral.
IMSI and time field to httplog data are extracted, because the two fields belong to httplog data
Elementary field, if there is arbitrary fields to be sky, then it is assumed that the data is imperfect, gives up the data.
2nd, the corresponding user's triple of httplog data is found;
2.1st, mobile phone MME data are obtained, IMSI, time and user's ternary group field is extracted;
Define 3:MME data;
MME (Mobility Management Entity) data are the crucial control sections of 3GPP agreement LTE access networks
Point, it is responsible for the UE (User Equipment) of idle pulley positioning, notifies from a phone call process, including relaying, briefly MME is negative
Blame signaling process part.There are IMSI, time and user's ternary group field in MME data, the corresponding relation of three can be set up.
Define 4:User's triple;
User's triple includes MmeUeS1apId, MmeGroupId and MmeCode, and the triple is distributed according to IMSI
Triple, as user's unique identifier, same IMSI within a period of time, user's triple of distribution be it is constant,
So corresponding user's triple can be found according to IMSI and time field.
2.2nd, the corresponding relation of IMSI- times tuple and user's triple is obtained.
Using IMSI and time field as a tuple, referred to as IMSI- time tuples, using the tuple as key values, are used
Family ternary group field is used as value values, you can set up IMSI- times tuple (IMSI, time) pass corresponding with user's triple
It is dictionary.It is now empty situation if there is IMSI, time or user's triple arbitrary fields, then needs to give up correspondence pass
System.
2.3rd, the IMSI by httplog data and time field, in the corresponding relation dictionary that the 2.2nd step is set up
Inquiry, obtains corresponding user's triple.Concretely comprise the following steps by the IMSI of httplog data and time field composition (IMSI, when
Between) tuple, using the tuple as key values, its value value is inquired about in the corresponding relation dictionary that the 2.2nd step is set up, successful inquiring is
Obtain corresponding user's triple.
3rd, the signal strength data corresponding to httplog data is obtained;
3.1st, mobile phone MR data are obtained, signal intensity and user's ternary group field is extracted;
Define 5:MR data;
MR (Measurement Report, measurement report) data refer to that information on a traffic channel (believe per 480ms by signaling
470ms on road) data are sent, these data can be used for network evaluation and optimization.The most important field of the data is signal
Intensity, can represent size of active user's mobile phone in the signal intensity at current time.MR data are stored with xml forms
, it is necessary to parsed to it, signal intensity and user's ternary group field can be extracted after parsing.
3.2nd, according to user's ternary group field of httplog data, in the corresponding signal intensity number of MR extracting datas
According to.
By the user's ternary group field inquired in the 2.3rd step, corresponding signal intensity can be extracted in MR data
Data.
4th, signal strength data is combined with httplog data, forms training data;
The label of training data is signal strength data, is characterized as the spy that other fields of httplog data are extracted
Levy.
5th, training data training pattern is used;
Will be by steps such as data statistics, feature extraction, model construction and model trainings during training pattern.
Data statistics, including field type statistics, Sparse degree statistics and missing Data-Statistics, while can also enter line number
According to steps such as cleaning and data conversions.
Feature extraction step, refers to that various dimensions carry out feature extraction at many levels, is found by data statistics in data
Rule, excavates valuable feature, in addition to the essential characteristic easily extracted, emphasis is in terms of space-time characteristic and content characteristic
Set about.
Model construction step according to input feature vector, it is necessary to select suitable model to be trained, because data are than sparse,
Therefore GBDT (Gradient Boosting Decision Tree) model is used.The thought of the model use integrated study, with
Decision tree, being capable of learning of nonlinear functions relation well as basic classification device.
Model training, is referred to and model is trained using training data, while carrying out arameter optimization.
6th, it is predicted using the model trained;
Test is obtained with after httplog data, by with the 5th step identical data statistics, feature extraction step, formed and surveyed
Data are tried as the input of model, the output result of model is to predict the outcome.Its data statistics and feature extraction side during prediction
It is similar when method is with training, but change over time, its model needs to constantly update, it is necessary to which constantly training new model is predicted.
The advantages of the present invention:
The present invention proposes a kind of method that cellular signal strength is predicted according to surfing Internet with cell phone daily record, and this method can be utilized
The huge signal strength data of the small httplog data prediction storage costs of storage cost, saves carrying cost, while being easy to fortune
Seek business and carry out the network optimization.Current operator can only store fraction signal strength data, therefore seldom be carried out using the data
Research.A large amount of signal strength datas can be predicted according to internet log using operator after the invention, can be in user behavior point
The research of deeper is carried out in terms of analysis, advertisement putting.
Brief description of the drawings
Fig. 1 is the flow chart that cellular signal strength method is predicted according to surfing Internet with cell phone daily record data.
Embodiment
The present invention is illustrated below in conjunction with embodiment.
1st step, the httplog data for obtaining all mobile phones of current period, extract its IMSI and time field
When httplog data refer to that user uses mobile network, the daily record data that base station is preserved.What the data were included
Information has user profile, base station information, behavioural information, temporal information.Wherein there are IMSI fields in user profile, be difference movement
The mark of user, is stored in SIM card, is the effective information for distinguishing mobile subscriber.IMSI to httplog data and when
Between field extracted because the two fields belong to the elementary field of httplog data, if having any one field for sky,
Think that the data is imperfect, give up the data.Table 1 is enumerated partial data field in httplog.
Table 1httplog data fields
Field | Explanation |
IMSI | Mobile phone card |
IMEI | Mobile phone string number |
MSISDN | Phone number |
USER_AGENT | Browser type |
APP_TYPE | Using major class |
APP_SUB_TYPE | Using group |
HOST | Access domain name |
title | Web site name |
keyword | Website keyword |
CELL_ID | Base station |
DL_DATA | Downlink traffic |
UL_DATA | Uplink traffic |
PROCEDURE_STARTTIME | Time started |
PROCEDURE_ENDTIME | End time |
PORTAL_APP_SET | Portal application set |
2nd step, find the corresponding user's triple of httplog data
User's triple includes MmeUeS1apId, MmeGroupId and MmeCode, and the triple is distributed according to IMSI
Triple, as user's unique identifier, same IMSI within a period of time, user's triple of distribution be it is constant,
So corresponding user's triple can be found according to IMSI and time field.IMSI- tempons can be obtained from MME data
The corresponding relation of group and user's triple.Using IMSI and time field as a tuple, using the tuple as key values, use
Family ternary group field is used as value values, you can set up the corresponding relation dictionary of IMSI- times tuple and user's triple.Now
It is empty situation if there is IMSI, time or user's triple arbitrary fields, then needs to give up the corresponding relation.
By the IMSI of httplog data, time field, inquire about, obtain in the corresponding relation dictionary set up before
Corresponding user's triple.Table 2 is enumerated part MME data fields.
Table 2MME data fields
Signal strength data corresponding to 3rd step, acquisition httplog data
First obtain mobile phone MR data, extract signal intensity, user's ternary group field, MR (Measurement Report,
Measurement report) data refer to information on a traffic channel per 480ms (470ms on signaling channel) send a data, these number
According to available for network evaluation and optimization.The most important field of the data is signal intensity, can represent that active user's mobile phone is being worked as
The size of the signal intensity at preceding moment.MR data be with xml forms stored, it is necessary to be parsed to it, after parsing
Signal intensity and user's ternary group field can be extracted.Table 3 is part MR data field explanations.
According to user's ternary group field of httplog data, corresponding signal strength data can be extracted in MR data.
Table 3MR data fields
4th step, signal strength data is combined with httplog data, forms training data
The label of training data is signal strength data, is characterized as the spy that other fields of httplog data are extracted
Levy.
5th step, use training data training pattern
Will be by steps such as data statistics, feature extraction, model construction and model trainings during training pattern.
Data statistics includes field type statistics, Sparse degree statistics, missing Data-Statistics, while can also carry out data
The steps such as cleaning, data conversion.
Feature extraction step refers to that various dimensions carry out feature extraction at many levels, and the rule in data are found by data statistics
Rule, excavates valuable feature, in addition to the essential characteristic easily extracted, and emphasis is in terms of space-time characteristic and content characteristic
Hand.
Model construction step needs to select suitable model to be trained according to input feature vector, because data are than sparse,
Therefore GBDT (Gradient Boosting Decision Tree) model is used.The thought of the model use integrated study, with
Decision tree, being capable of learning of nonlinear functions relation well as basic classification device.
Model training refers to be trained using training data to model, while carrying out arameter optimization.It is special that table 4 lists part
Levy.
The Partial Feature of table 4
Feature | Description |
download_flux | Downlink traffic number |
total_flux | Total flow number |
app_type | The application major class that user uses |
cell_area | Base station covering scene |
time_stamp | Temporal characteristics |
buss_behavior_flag | Current business behavior |
buss_finish_flag | Business completes mark |
buss_browser | Operational Visit instrument |
http_content | Http transferring contents |
6th step, it is predicted using the model trained
Test is obtained with after httplog data, by with the 5th step identical data statistics, feature extraction step, formed and surveyed
Data are tried as the input of model, the output result of model is to predict the outcome.Wherein, the part sample of feature extraction such as table 4
It is shown.The model used is GBDT models, and as shown in table 5, other specification uses model default parameters to its parameter.Use test number
According to as mode input, while the parameter of allocation list 5, can obtain the output result of model, as finally predict the outcome.Prediction knot
Fruit sample and as shown in table 6 with the contrast of training examples.Its phase difference is little, therefore all true that can not obtain signal intensity
In the case of real value, it is feasible to substitute actual value to carry out other schemes studied using predicted value.Its data statistics during prediction
It is similar during with Feature Extraction Method with training, but change over time, its model needs to constantly update, it is necessary to constantly train new mould
Type is predicted.
Table 5GBDT configuration parameters
Parameter | Arranges value | Description |
max_depth | 15 | Depth capacity |
min_child_weight | 0.5 | The minimal weight of child node |
eta | 0.1 | Iteration step length |
lambda | 1 | L2 regularization term weights |
objective | reg:logistic | Task type |
eval_metric | rmse | Evaluation criterion |
Table 6 predicts the outcome sample
Claims (3)
1. a kind of method that cellular signal strength is predicted according to surfing Internet with cell phone daily record, its step is:
1st, the httplog data of all mobile phones of current period are obtained, IMSI and time field is extracted;
Define 1:Httplog data;It is defined as follows:
When httplog data refer to that user uses mobile network, the daily record data that base station is preserved;The information that the data are included
Mainly have:
(1) user profile:Mainly include the essential information of user mobile phone;
(2) base station information:The main essential information that base station is connected including user;
(3) behavioural information:Refer to the specific internet behavior of user;
(4) temporal information:When referring to user's generation internet behavior, the information relevant with the time;
Define 2:IMSI fields;It is defined as follows:
IMSI fields refer to international mobile subscriber identity (International Mobile Subscriber
Identification Number), it is the mark for distinguishing mobile subscriber, is stored in SIM card, is to be used to distinguish mobile subscriber
Effective information;IMSI fields total length be no more than 15, equally using 0~9 numeral;
2nd, the corresponding user's triple of httplog data is found;
2.1st, mobile phone MME data are obtained, IMSI, time and user's ternary group field is extracted;
Define 3:MME data;It is defined as follows:
MME (Mobility Management Entity) is the key control node of 3GPP agreement LTE access networks, and it is responsible for
The UE (User Equipment) of idle pulley positioning, notifies from a phone call process, including relaying, briefly MME is responsible at signaling
Manage part;There are IMSI, time and user's ternary group field in MME data, the corresponding relation of three can be set up;
Define 4:User's triple;It is defined as follows:
User's triple includes MmeUeS1apId, MmeGroupId and MmeCode, and the triple is three distributed according to IMSI
Tuple, as user's unique identifier, same IMSI is within a period of time, and user's triple of distribution is constant, so
Corresponding user's triple can be found according to IMSI and time field;
2.2nd, the corresponding relation of IMSI- times tuple and user's triple is obtained;It regard IMSI and time field as a member
Group, referred to as IMSI- time tuples, using the tuple as key values, user's ternary group field is used as value values, you can set up
The corresponding relation dictionary of IMSI- times tuple (IMSI, time) and user's triple;
2.3rd, the IMSI by httplog data and time field, are looked into the corresponding relation dictionary that the 2.2nd step is set up
Ask, obtain corresponding user's triple;
3rd, the signal strength data corresponding to httplog data is obtained;
3.1st, mobile phone MR data are obtained, signal intensity and user's ternary group field is extracted;
Define 5:MR data;It is defined as follows:
MR (Measurement Report, measurement report) data refer to information on a traffic channel per 480ms or signaling channel
Upper 470ms sends a data, and these data can be used in network evaluation and optimization;The most important field of the data is that signal is strong
Degree, can represent size of active user's mobile phone in the signal intensity at current time;
MR data be with xml forms stored, it is necessary to parsed, signal intensity and user three can be extracted after parsing
Tuple field;
3.2nd, according to user's ternary group field of httplog data, in the corresponding signal strength data of MR extracting datas;It is logical
The user's ternary group field inquired in the 2.3rd step is crossed, can be in the corresponding signal strength data of MR extracting datas;
4th, signal strength data is combined with httplog data, forms training data;
5th, training data training pattern is used;
6th, it is predicted using the model trained.
2. according to the method described in claim 1, it is characterised in that described in the 5th step during training pattern, by data statistics,
Feature extraction, model construction and model training step, concrete function are as follows:
(1) data statistics:Including field type statistics, Sparse degree statistics and missing Data-Statistics, while can also enter line number
According to cleaning and data conversion step, the feature extraction after convenience works;
(2) feature extraction:Various dimensions carry out feature extraction at many levels, and the rule in data is found by data statistics, excavate
Valuable feature, in addition to the essential characteristic easily extracted, emphasis is set about in terms of space-time characteristic and content characteristic;
(3) model construction:Select suitable model to be trained according to input feature vector, because data are than sparse, therefore use
GBDT (Gradient Boosting Decision Tree) model;The thought of the model use integrated study, is made with decision tree
, being capable of learning of nonlinear functions relation well for basic classification device;
(4) model training:Model is trained using training data, while carrying out arameter optimization.
3. according to the method described in claim 1, it is characterised in that data statistics and feature when being predicted described in the 6th step
It is similar when abstracting method is with training, but change over time, model needs are constantly updated, it is necessary to which constantly training new model is carried out in advance
Survey.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710544918.6A CN107333285B (en) | 2017-07-06 | 2017-07-06 | Method for predicting mobile phone signal strength according to mobile phone internet log |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710544918.6A CN107333285B (en) | 2017-07-06 | 2017-07-06 | Method for predicting mobile phone signal strength according to mobile phone internet log |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107333285A true CN107333285A (en) | 2017-11-07 |
CN107333285B CN107333285B (en) | 2020-09-18 |
Family
ID=60195840
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710544918.6A Active CN107333285B (en) | 2017-07-06 | 2017-07-06 | Method for predicting mobile phone signal strength according to mobile phone internet log |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107333285B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113792019A (en) * | 2021-08-03 | 2021-12-14 | RealMe重庆移动通信有限公司 | Analysis method, electronic device and computer storage medium |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102681886A (en) * | 2011-04-14 | 2012-09-19 | 天脉聚源(北京)传媒科技有限公司 | Method and system for tracking user behaviors on mobile equipment |
CN105451348A (en) * | 2014-09-25 | 2016-03-30 | 中兴通讯股份有限公司 | Network control method and device |
CN105792138A (en) * | 2016-02-26 | 2016-07-20 | 重庆邮电大学 | Indoor WLAN user positioning and behavior analysis method based on line segment density |
US20160269984A1 (en) * | 2013-12-12 | 2016-09-15 | Locality Systems Inc. | Proximity recognition system |
CN106332138A (en) * | 2016-08-31 | 2017-01-11 | 四川亨通网智科技有限公司 | Method for obtaining MRO user information based on S1-MME signalling association |
CN106535234A (en) * | 2016-12-31 | 2017-03-22 | 中国移动通信集团江苏有限公司 | Method and device for associating measurement report with user, and method and device for locating Internet surfing problem of user |
CN106804046A (en) * | 2017-02-16 | 2017-06-06 | 广州杰赛科技股份有限公司 | Mobile location method and device based on measurement report |
CN106900004A (en) * | 2017-03-14 | 2017-06-27 | 上海奕行信息科技有限公司 | The method of association LTE network wireless measurement report MR record ownership IMSI numbers |
CN106912061A (en) * | 2017-02-28 | 2017-06-30 | 中国联合网络通信集团有限公司 | A kind of mobile terminal limit accesses the computational methods and device of distance |
-
2017
- 2017-07-06 CN CN201710544918.6A patent/CN107333285B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102681886A (en) * | 2011-04-14 | 2012-09-19 | 天脉聚源(北京)传媒科技有限公司 | Method and system for tracking user behaviors on mobile equipment |
US20160269984A1 (en) * | 2013-12-12 | 2016-09-15 | Locality Systems Inc. | Proximity recognition system |
CN105451348A (en) * | 2014-09-25 | 2016-03-30 | 中兴通讯股份有限公司 | Network control method and device |
CN105792138A (en) * | 2016-02-26 | 2016-07-20 | 重庆邮电大学 | Indoor WLAN user positioning and behavior analysis method based on line segment density |
CN106332138A (en) * | 2016-08-31 | 2017-01-11 | 四川亨通网智科技有限公司 | Method for obtaining MRO user information based on S1-MME signalling association |
CN106535234A (en) * | 2016-12-31 | 2017-03-22 | 中国移动通信集团江苏有限公司 | Method and device for associating measurement report with user, and method and device for locating Internet surfing problem of user |
CN106804046A (en) * | 2017-02-16 | 2017-06-06 | 广州杰赛科技股份有限公司 | Mobile location method and device based on measurement report |
CN106912061A (en) * | 2017-02-28 | 2017-06-30 | 中国联合网络通信集团有限公司 | A kind of mobile terminal limit accesses the computational methods and device of distance |
CN106900004A (en) * | 2017-03-14 | 2017-06-27 | 上海奕行信息科技有限公司 | The method of association LTE network wireless measurement report MR record ownership IMSI numbers |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113792019A (en) * | 2021-08-03 | 2021-12-14 | RealMe重庆移动通信有限公司 | Analysis method, electronic device and computer storage medium |
CN113792019B (en) * | 2021-08-03 | 2023-08-18 | RealMe重庆移动通信有限公司 | Analysis method, electronic equipment and computer storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN107333285B (en) | 2020-09-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10477416B2 (en) | Network traffic forecasting for non-ticketed events | |
CN106997357B (en) | Message processing method, device and system | |
CN104796481B (en) | Intelligent audio and video selection method | |
CN105007171A (en) | User data analysis system and method based on big data in communication field | |
CN105306495B (en) | user identification method and device | |
CN105631707A (en) | Advertisement click rate estimation method based on decision tree, application recommendation method and device | |
CN103258022A (en) | Local commerce service recommendation system and method based on user interest | |
CN104331404A (en) | A user behavior predicting method and device based on net surfing data of a user's cell phone | |
CN103970861A (en) | Information presenting method and device | |
CN110049079A (en) | Information push and model training method, device, equipment and storage medium | |
CN113412608B (en) | Content pushing method and device, server and storage medium | |
US20160210647A1 (en) | Method of determining segmentations of subscribers, network entity using the same, and server using the same | |
CN108462965B (en) | Iron tower demand site sharing planning method | |
CN103390000A (en) | Web searching method and web searching system | |
CN104750760A (en) | Application software recommending method and device | |
CN102999513B (en) | Based on information displaying method and the device of geographic position service search | |
WO2014183544A1 (en) | Method and device for generating a personalized navigation webpage | |
CN109743745A (en) | The recognition methods of mobile network's access style, device, electronic equipment and storage medium | |
CN105095370A (en) | Information push method, device and system for expansion application | |
CN108205770A (en) | The acquisition methods and device of a kind of target medium | |
Li et al. | Dissecting foursquare venue popularity via random region sampling | |
WO2015196377A1 (en) | Method and device for determining user identity category | |
CN107333285A (en) | A kind of method that cellular signal strength is predicted according to surfing Internet with cell phone daily record | |
CN105578395B (en) | A kind of method and device updating terminal attribute in end message library | |
CN105491136B (en) | Message method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |