CN106685721B - method and system for calculating predictability of user online activity outbreak time - Google Patents
method and system for calculating predictability of user online activity outbreak time Download PDFInfo
- Publication number
- CN106685721B CN106685721B CN201611264520.9A CN201611264520A CN106685721B CN 106685721 B CN106685721 B CN 106685721B CN 201611264520 A CN201611264520 A CN 201611264520A CN 106685721 B CN106685721 B CN 106685721B
- Authority
- CN
- China
- Prior art keywords
- user
- activity
- time
- time interval
- burst
- 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.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method and a system for calculating the predictability of the user online activity outbreak time, wherein the method comprises the following steps: extracting a burst cluster structure from a historical online activity time record of a user; acquiring a user outbreak time interval sequence by utilizing an outbreak cluster structure; discretizing the burst time interval sequence of the user to obtain a burst time interval symbol sequence; calculating the entropy rate of a user burst time interval symbol sequence; and calculating the degree of predictability of the user online activity outbreak time according to the entropy rate of the user outbreak time interval symbol sequence. The invention provides a method for calculating the predictability of the online activity outbreak time of a user, which can better predict the next online activity behavior of the user and help developers to better design and improve an online service platform.
Description
Technical Field
The invention relates to the technical field of internet, in particular to a method for calculating the predictability of the user online activity outbreak time.
Background
With the rapid development of the internet, particularly the mobile internet technology, many activities that people need to complete in real life can be selected to be performed on the network in the past, and meanwhile, the internet technology also provides more and more new services and applications for the daily life of people. The widespread use of internet technology has made online activities play an increasingly important role in people's daily life. For human behavior on the internet, namely, online behavior of a user, knowing next activity of the user in advance can help developers to better design and improve an online service platform.
Compared with the design of the user online behavior prediction algorithm, a more basic problem is to obtain a prediction performance bound of the user online behavior, namely an upper bound of the prediction accuracy rate which can be achieved by the prediction algorithm. The calculation of the upper bound of the prediction accuracy can guide the design of a prediction algorithm and help the comprehensive analysis algorithm to optimize space and research cost control.
the user's activities on the internet present an explosive feature, i.e. a large number of activities can be generated in a short time after a long period of inactivity. The outbreak time of the user online activity is a key dimension for describing the user online behavior, and no calculation method for the predictability of the user online activity outbreak time exists at present.
disclosure of Invention
The invention mainly aims to provide a method for calculating the predictability of the online activity outbreak time of a user.
The invention provides a method for calculating the predictability of the user online activity outbreak time, which comprises the following steps:
extracting a burst cluster structure from a historical online activity time record of a user;
Acquiring a user outbreak time interval sequence by utilizing an outbreak cluster structure;
Discretizing the burst time interval sequence of the user to obtain a burst time interval symbol sequence;
Calculating the entropy rate of a user burst time interval symbol sequence;
and calculating the degree of predictability of the user online activity outbreak time according to the entropy rate of the user outbreak time interval symbol sequence.
Further, before extracting the burst cluster structure from the historical online activity time record of the user, including,
and acquiring all activity time records of the user on a specified platform.
Further, the step of extracting the burst cluster structure from the historical online activity time record of the user comprises,
And setting a time interval threshold of related activities, and dividing all activity time records of the user into a burst cluster structure.
Further, the step of setting the time interval threshold of the related activities and dividing all activity time records of the user into burst cluster structures comprises,
setting a relative activity time interval threshold, judging whether the time interval of two activities is less than the threshold or not,
if so, judging that the two activities are related, and taking the two activities as the same outbreak cluster;
if not, it is determined that the two activities are not related.
further, the step of setting the time interval threshold of the related activities and dividing all activity time records of the user into burst cluster structures further comprises,
Setting a first activity time record of a user as a first activity in a first outbreak cluster, starting from a second activity time record of the user, judging one by one, if the time interval between the first activity time record and the previous activity is smaller than a set related activity time interval threshold, dividing the first activity time record into the cluster where the previous activity record is located, if the time interval between the first activity time record and the previous activity is larger than the set related activity time interval threshold, taking the first activity time record as the first activity of a new outbreak activity cluster, and marking the first activity time record as the starting time of the outbreak activity cluster.
Further, the discretizing of the burst interval sequence of the user to obtain a burst interval symbol sequence includes,
and discretizing by adopting an equal frequency discretization method.
Further, the discretizing step by using the equal-frequency discretization method includes:
after all the explosion time interval sequences of the user are obtained, firstly, a discrete interval is divided, each explosion time interval of the user is placed in the corresponding discrete interval, the placing rule is that the explosion time interval is larger than or equal to the left end point value of the placed interval and is smaller than the right end point value of the placed interval, and after one explosion time interval is placed, the serial number of the corresponding discrete interval is the discretization result corresponding to the explosion time interval.
Further, after obtaining all the burst time interval sequences of the user, the step of dividing a discrete interval first comprises,
dividing k discrete intervals between the maximum value and the minimum value of the burst time interval, wherein the division of the intervals ensures that the frequency of all burst time intervals falling into each interval is equal, and the 1 st, 2 nd, … th and k total k discrete intervals which are arranged from small to large can be obtained according to the method.
A user online activity break out time predictability calculation system, comprising:
and the extraction unit is used for extracting the burst cluster structure from the historical online activity time record of the user.
an obtaining unit, configured to obtain a user burst interval sequence by using a burst cluster structure.
and the discrete unit is used for discretizing the burst time interval sequence of the user to obtain a burst time interval symbol sequence.
And the calculating unit is used for calculating the entropy rate of the user explosion time interval symbol sequence and calculating the degree of predictability of the user online activity explosion time according to the entropy rate of the user explosion time interval symbol sequence.
further, the extraction unit includes,
a dividing module: and setting a time interval threshold of related activities, and dividing all activity time records of the user into a burst cluster structure.
The invention has the beneficial effects that: the method can better predict the next online activity behavior of the user, and helps developers to better design and improve an online service platform.
Drawings
FIG. 1 is a flow chart of a method for calculating the predictability of the burst time of a user's online activity according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method for calculating the predictability of the burst time of a user's online activity according to another embodiment of the invention;
Fig. 3 is a block diagram of a predictive online activity burst time calculation system based on internet user data according to another embodiment of the present invention.
FIG. 4 is a time interval threshold diagram illustrating a method for calculating the predictability of the burst time of user online activity according to an embodiment of the present invention;
Fig. 5 is a time interval sequence diagram of a method for calculating the predictability of the burst time of the user's online activity according to an embodiment of the present invention.
the implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
referring to fig. 1, a method for calculating the predictability of the burst time of a user's online activity includes the steps of:
S10, extracting a burst cluster structure from the historical online activity time record of the user;
S11, acquiring a user explosion time interval sequence by utilizing an explosion cluster structure;
S12, discretizing the burst time interval sequence of the user to obtain a burst time interval symbol sequence;
S13, calculating the entropy rate of the user burst time interval symbol sequence;
and S14, calculating the user online activity burst time predictability according to the entropy rate of the user burst time interval symbol sequence.
In the step S10, the burst cluster structure can better reflect the activity of the user in the similar time, and provide an important basis for the step S11 to obtain the time interval sequence and distinguish the time periods.
In the step S12, the discretization is to map infinite individuals in infinite space into finite space to improve the space-time efficiency of the algorithm, and in this step, the time interval sequence is discretized to obtain a time interval symbol sequence.
in the above step S13, entropy rate refers to the average uncertainty of a random source (a random process that continuously generates random variables) over time. The entropy rate of a random process is the uncertainty caused by the average generation of each random variable in the process, and after the time interval symbol sequence is obtained in step S12, according to the time interval symbol sequence, the formula is calculated:
calculating an entropy rate of a sequence of time interval symbols, wherein SestRepresenting the entropy rate of the user burst interval symbol sequence, n representing the length of the user burst interval symbol sequence, table Λhthe length of the shortest sub-sequence starting from the h-th symbol in the sequence and not appearing in the sub-columns formed by the 1 st to the h-1 st symbols is shown.
in step S14, the entropy rate of the time interval symbol sequence is calculated as the degree of predictability of the user burst time interval symbol sequence, and the formula is calculated as follows:
Sest＝-[Πlog2Π+(1-Π)log2(1-Π)]+(1-Π)log2(N-1)
wherein pi represents the predictability of the user explosion time interval symbol sequence, namely the maximum value of the prediction accuracy rate which can be reached by predicting the user explosion time, N represents the number of different symbols in the user explosion time interval symbol sequence, and the result of the predictability of the explosion time provided by the invention is between 0 and 100 percent.
in this embodiment, an outbreak cluster structure is obtained according to a part of a user on a certain platform or a certain activity time, a time interval sequence is obtained according to the outbreak cluster structure, the time interval sequence is discretized to obtain a time interval symbol sequence, and finally the predictability of the user's online activity outbreak time is calculated by using the information entropy and the feinuo inequality.
in an embodiment of the present invention, the burst cluster refers to a series of high frequency activities that are generated in a burst manner in a short time during the activity of the human being.
In an embodiment of the present invention, the burst interval sequence refers to an interval sequence composed of intervals of start timings of adjacent burst clusters.
In an embodiment of the present invention, the burst interval symbol sequence refers to a symbol sequence obtained by performing symbol conversion on a burst interval sequence by using a corresponding relationship between a burst interval and a burst interval symbol provided by a discretization method.
referring to fig. 2, in another embodiment of the present invention, a method for calculating the predictability of the burst time of user online activity comprises the following steps:
S20, acquiring all activity time records of the user on a certain platform;
S21, setting a time interval threshold of related activities, and dividing all activity time records of the user into outbreak cluster structures;
S22, acquiring a user explosion time interval sequence by utilizing an explosion cluster structure;
S23, discretizing the burst time interval sequence of the user by adopting an equal-frequency discretization method to obtain a burst time interval symbol sequence;
S24, calculating the entropy rate of the user burst time interval symbol sequence;
and S25, calculating the user online activity burst time predictability according to the entropy rate of the user burst time interval symbol sequence.
In the step S21, the method for setting the time interval threshold of the relevant activities includes:
two activities are considered related when the time interval between the activities is less than the threshold, and are considered unrelated when the time interval between the activities is greater than the threshold.
in the above step S21, the method for dividing all activity time records of the user into burst cluster structures includes:
Setting a first activity time record of a user as a first activity in a first outbreak cluster, starting from a second activity time record of the user, judging one by one, if the time interval between the first activity time record and the previous activity is smaller than a set related activity time interval threshold, dividing the first activity time record into the cluster where the previous activity record is located, if the time interval between the first activity time record and the previous activity is larger than the set related activity time interval threshold, taking the first activity time record as the first activity of a new outbreak activity cluster, and marking the first activity time record as the starting time of the outbreak activity cluster.
In the embodiment of the present invention, discretizing a burst interval sequence of a user to obtain a burst interval symbol sequence includes:
in this embodiment, in the step S23, an equal-frequency discretization method is adopted for discretization:
Dividing a discrete interval after obtaining all the explosion time interval sequences of the user, placing each explosion time interval of the user in the corresponding discrete interval, wherein the placing rule is that the explosion time interval is larger than or equal to the left end point value of the placed interval and is smaller than the right end point value of the placed interval, and after placing one explosion time interval, the serial number of the corresponding interval is the discretization result corresponding to the explosion time interval. The method for dividing the discrete intervals comprises the following steps: dividing k discrete intervals between the maximum value and the minimum value of the burst time interval, wherein the division of the intervals ensures that the frequency of all burst time intervals falling into each interval is equal, and the 1 st, 2 nd, … th and k total k discrete intervals which are arranged from small to large can be obtained according to the method.
in the embodiment of the invention, when the predictability of the outbreak activity time of a certain user on a certain network platform needs to be calculated, all activity time records of the user on the platform are obtained firstly. After obtaining the activity time record, firstly setting a related activity time interval threshold which is used for distinguishing whether two activities are related or not, and when the time interval of the two activities is smaller than the threshold, considering the two activities to be related, namely the two activities are from the same outbreak cluster; when the time interval between two activities is greater than the threshold, the two activities are considered to be irrelevant, that is, the two activities originate from two different outbreak clusters, and the value of the relevant activity time interval threshold is related to a specific network platform.
And dividing all activity time records of the user on the platform into burst cluster structures according to the set related activity time interval threshold. For the first activity time record of the user, it is set to the first activity in the first burst cluster. Starting from the second activity time record of the user, judging one by one, and if the time interval between the second activity time record and the previous activity is smaller than the set related activity time interval threshold, dividing the second activity time record into the cluster where the previous activity record is located; if the time interval between the current activity and the previous activity is larger than the set related activity time interval threshold, the current activity is taken as the first activity of a new burst activity cluster, and the current activity is marked as the starting time of the burst activity cluster. As shown in fig. 4, Δ t represents the relative activity time interval threshold.
After dividing all activity records of the user into the burst cluster structure, the start times of all burst clusters of the user on the platform can be obtained, and the burst activity time interval sequence of the user can be obtained from the burst start time of each cluster, as shown in fig. 5, tau1，τ2，τ3，τ4Representing a portion of a user's burst activity interval sequence.
after the outbreak activity time interval sequence of the user is obtained, the invention uses a discretization method to discretize the outbreak activity time interval to obtain an outbreak activity time interval symbol sequence, and the predictability is calculated through the time interval symbol sequence.
Because the burst time interval distribution of the user has a fat tail characteristic, in order to realize effective burst time interval symbolization, the invention adopts an equal-frequency discretization method to carry out discretization, and after all burst time interval sequences of the user are obtained, the maximum value and the minimum value of the burst time interval are obtained. K discrete intervals are divided between the maximum and minimum values, the division of the intervals being such that the frequency with which all burst intervals fall within each interval is equal. According to the method, k discrete intervals which are arranged from small to large and are 1, 2, … and k are obtained.
After all the discrete intervals are obtained, each burst time interval of the user is placed in the corresponding interval according to the rule that the burst time interval is larger than or equal to the left end point value of the put-in interval and smaller than the right end point value of the put-in interval. After a burst time interval is placed, the serial number of the corresponding interval is the discretization result corresponding to the burst time interval, so that each burst time interval corresponds to a burst time interval symbol, the time interval symbols are 1 ', 2 ', … and k ', and after each time interval in the user burst time interval sequence is discretized, the burst time interval sequence of the user can be converted into a burst time interval symbol sequence.
After the outbreak time interval symbol sequence of the user is obtained, the method utilizes the information entropy and the Voronoi inequality to calculate the degree of predictability of the outbreak time of the online activity of the user. The calculation is as follows.
firstly, calculating the entropy rate of a user burst time interval symbol sequence, wherein the calculation formula is as follows:
wherein S isestrepresenting the entropy rate of the user burst interval symbol sequence, n representing the length of the user burst interval symbol sequence, table ΛhThe length of the shortest sub-sequence starting from the h-th symbol in the sequence and not appearing in the sub-column formed by the 1 st to the h-1 st symbols is shown.
After the user outbreak time interval sequence is obtained through calculation, the degree of predictability of the corresponding symbol sequence is calculated through the Voronoi inequality, namely, the maximum value of the prediction accuracy rate which can be reached by a prediction algorithm when the next symbol is predicted through all historical symbols in the sequence is utilized. The calculation formula is as follows.
Sest＝-[Πlog2Π+(1-Π)log2(1-Π)]+(1-Π)log2(N-1)
Wherein Π represents the degree of predictability of the user explosion time interval symbol sequence, that is, the maximum value of the prediction accuracy rate that the user explosion time can reach, and N represents the number of different symbols in the user explosion time interval symbol sequence.
Referring to fig. 3, the present invention also provides an online activity outbreak time predictability calculation system based on internet user data, comprising:
An extracting unit 100, configured to extract a burst cluster structure from a historical online activity time record of a user.
an obtaining unit 110, configured to obtain a user burst interval sequence by using a burst cluster structure.
the discretization unit 120 is configured to discretize the burst interval sequence of the user to obtain a burst interval symbol sequence.
And the calculating unit 130 is configured to calculate an entropy rate of the user burst interval symbol sequence, and calculate the user online activity burst time predictability according to the entropy rate of the user burst interval symbol sequence.
In another embodiment, the extracting unit is configured to extract the burst cluster structure from the historical online activity time record of the user, and includes,
the dividing module 1001: and setting a time interval threshold of related activities, and dividing all activity time records of the user into a burst cluster structure.
the above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (7)
1. a method for calculating the predictability of the burst time of a user's online activity, comprising the steps of:
Acquiring all activity time records of the user on a specified platform, extracting an outbreak cluster structure from historical online activity time records of the user, setting a time interval threshold of related activities, and dividing all activity time records of the user into the outbreak cluster structure;
acquiring a user outbreak time interval sequence by utilizing an outbreak cluster structure;
Discretizing the burst time interval sequence of the user to obtain a burst time interval symbol sequence;
calculating the entropy rate of a user burst time interval symbol sequence;
and calculating the degree of predictability of the user online activity outbreak time according to the entropy rate of the user outbreak time interval symbol sequence.
2. The method for calculating the predictability of the burst time of the user's online activity according to claim 1, wherein said step of setting the time interval threshold of the related activity, dividing all activity time records of the user into burst cluster structure comprises,
setting a relative activity time interval threshold, judging whether the time interval of two activities is less than the threshold or not,
If so, judging that the two activities are related, and taking the two activities as the same outbreak cluster;
if not, it is determined that the two activities are not related.
3. The method for calculating predictability of user online activity burst time according to claim 2, wherein said step of setting a time interval threshold of related activities, dividing all activity time records of the user into burst cluster structure, further comprises,
Setting a first activity time record of a user as a first activity in a first outbreak cluster, starting from a second activity time record of the user, judging one by one, if the time interval between the first activity time record and the previous activity is smaller than a set related activity time interval threshold, dividing the first activity time record into the cluster where the previous activity record is located, if the time interval between the first activity time record and the previous activity is larger than the set related activity time interval threshold, taking the first activity time record as the first activity of a new outbreak activity cluster, and marking the first activity time record as the starting time of the outbreak activity cluster.
4. The method for calculating the predictability of the burst time of the user's online activity according to claim 1, wherein the discretizing the sequence of burst time intervals of the user to obtain a sequence of burst time interval symbols comprises,
And discretizing by adopting an equal frequency discretization method.
5. The method for calculating the predictability of the user online activity burst time according to claim 4, wherein the discretizing step by using the constant frequency discretization method comprises the following steps:
after all the explosion time interval sequences of the user are obtained, firstly, a discrete interval is divided, each explosion time interval of the user is placed in the corresponding discrete interval, the placing rule is that the explosion time interval is larger than or equal to the left end point value of the placed interval and is smaller than the right end point value of the placed interval, and after one explosion time interval is placed, the serial number of the corresponding discrete interval is the discretization result corresponding to the explosion time interval.
6. the method for calculating the predictability of the burst time of the user's online activity according to claim 5, wherein said step of dividing a discrete interval after obtaining all the burst time interval sequences of the user comprises,
dividing k discrete intervals between the maximum value and the minimum value of the burst time interval, wherein the division of the intervals ensures that the frequency of all burst time intervals falling into each interval is equal, and the 1 st, 2 nd, … th and k total k discrete intervals which are arranged from small to large can be obtained according to the method.
7. A user online activity break out time predictability calculation system, comprising:
The extraction unit is used for acquiring all activity time records of the user on a specified platform and extracting an outbreak cluster structure from the historical online activity time records of the user; and a partitioning module within the extraction unit: setting a time interval threshold of related activities, and dividing all activity time records of a user into outbreak cluster structures;
an acquisition unit, configured to acquire a user burst interval sequence using a burst cluster structure;
the discrete unit is used for discretizing the burst time interval sequence of the user to obtain a burst time interval symbol sequence;
and the calculating unit is used for calculating the entropy rate of the user explosion time interval symbol sequence and calculating the degree of predictability of the user online activity explosion time according to the entropy rate of the user explosion time interval symbol sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611264520.9A CN106685721B (en) | 2016-12-30 | 2016-12-30 | method and system for calculating predictability of user online activity outbreak time |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611264520.9A CN106685721B (en) | 2016-12-30 | 2016-12-30 | method and system for calculating predictability of user online activity outbreak time |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106685721A CN106685721A (en) | 2017-05-17 |
| CN106685721B true CN106685721B (en) | 2019-12-13 |
Family
ID=58850603
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611264520.9A Active CN106685721B (en) | 2016-12-30 | 2016-12-30 | method and system for calculating predictability of user online activity outbreak time |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106685721B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107122767B (en) * | 2017-05-27 | 2020-04-10 | 辽宁大学 | Human activity endpoint detection method based on information entropy |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102118245A (en) * | 2009-12-31 | 2011-07-06 | 中国人民解放军国防科学技术大学 | Scale prediction knowledge training method and prediction method for large-scale network security events |
| CN104679992A (en) * | 2015-01-30 | 2015-06-03 | 南京邮电大学 | Markov model designing method based on user service use time |
-
2016
- 2016-12-30 CN CN201611264520.9A patent/CN106685721B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102118245A (en) * | 2009-12-31 | 2011-07-06 | 中国人民解放军国防科学技术大学 | Scale prediction knowledge training method and prediction method for large-scale network security events |
| CN104679992A (en) * | 2015-01-30 | 2015-06-03 | 南京邮电大学 | Markov model designing method based on user service use time |
Non-Patent Citations (1)
| Title |
|---|
| "Specific Differential Entropy Rate Estimation for";Raúl Alcaraz Martínez;《entropy》;20160519;第1-3页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106685721A (en) | 2017-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lu et al. | Bias correction in a small sample from big data | |
| CN105302903A (en) | Search method, apparatus and system and search result sequence adjustment basis determination method | |
| CN103473036A (en) | Input method skin push method and system | |
| CN111061837A (en) | Topic identification method, device, equipment and medium | |
| CN106685721B (en) | method and system for calculating predictability of user online activity outbreak time | |
| CN106250499B (en) | Video pair mining method and device | |
| CN103957537A (en) | Frequency spectrum perception method and apparatus based on channel cluster | |
| CN111832150B (en) | Ore pressure prediction method and device and electronic equipment | |
| CN104133918B (en) | A kind of acquisition methods and device, method for pushing and device of interest point information | |
| JP5973935B2 (en) | Browsing behavior prediction device, browsing behavior prediction method, and program | |
| CN109657071B (en) | Vocabulary prediction method, device, equipment and computer readable storage medium | |
| US11859467B2 (en) | Reservoir simulation systems and methods to dynamically improve performance of reservoir simulations | |
| CN104778202A (en) | Analysis method and system based on event evolution process of key words | |
| CN114896451A (en) | Video clip positioning method, system, control device and readable storage medium | |
| KR101804020B1 (en) | Method for sns bot detection using geographic information | |
| CN113177418A (en) | Session theme determining method and device, storage medium and electronic equipment | |
| Liu et al. | HMM-based state prediction for Internet hot topic | |
| Yazji et al. | Protecting private data on mobile systems based on spatio–temporal analysis | |
| Adi et al. | The effect of data acquisition techniques in profiling analysis based on twitter | |
| CN110324292B (en) | Authentication device, authentication method, and computer storage medium | |
| RU107613U1 (en) | DISCRETE SIGNAL RECEIVER | |
| Blanc et al. | Quantifying neural correlations using lempel-ziv complexity | |
| CN112667610A (en) | Real-time processing method and device for internet of things data and computer equipment | |
| Deusdado et al. | Employing compact intra-genomic language models to predict genomic sequences and characterize their entropy | |
| Lou et al. | Traffic session identification based on statistical language model |
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 |