KR20170129219A - Method and apparatus for determining user demand - Google Patents

Abstract

A method and apparatus for determining user demand, the method comprising: querying (S101) a user search association rule based on a corresponding time sequence according to a user's past search history; And determining the demand information of the user according to the user search association rule (S102). This method can help the user to find the information he needs more conveniently and quickly.

Description

Method and apparatus for determining user demand

This application claims the priority of the Chinese patent application No. 201510659433.2, entitled "Method and Apparatus for Determining User Demand," filed on October 12, 2015, the contents of which are incorporated herein by reference .

The present invention relates to the field of electronic information, and more particularly to a method and apparatus for determining user demand.

As e-commerce and mobile communication networks become popular, the Internet is getting closer and closer to everyday life, and people are increasingly searching for various types of information on the Internet to purchase various goods and services. Therefore, it is a big problem how to enable users to acquire information required by each user more quickly. However, in the conventional technology, the web page always displays the goods or services that have been searched for in the past, , The web page displays some shoe or shoe stores to the user after the user purchases the shoes, but in most cases there is no longer a need for the user to purchase more shoes, which does not help the user , The user's time is wasted due to complicated and redundant information.

It is an object of the present invention to provide a method and an apparatus for assisting a user to determine a user demand that can find information required by each user more conveniently and quickly.

In one aspect of the present invention, there is provided a method of determining user demand,

Querying a user search association rule based on a time sequence based on a user's past search history; And

And determining the demand information of the user according to the user search association rule.

According to another aspect of the present invention, there is further provided an apparatus for determining demand information of a user,

An apparatus configured to query a user search association rule based on a time sequence based on a past search history of a user; And

And a device configured to determine demand information of the user according to a user search association rule.

Compared with the prior art, the present invention has the following advantages. The present invention determines a user's demand using a time sequence (or dynamic sequence) -based association rule. A time sequence is a sequence formed by arranging numerical values of the same statistical index in a temporal order of occurrence of a corresponding occurrence, The main purpose of the analysis is to predict the future based on existing historical data. Specifically, when people actually search for goods and services, because the demand for information often has a temporal order in time, the technical idea is based on time sequence, and when a user browses a web page, Information related to some past search records can be pushed forward in accordance with a past search history of the user and a predetermined association rule so that users can find information they need more conveniently and quickly.

Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments, which proceeds with reference to the accompanying drawings.
1 is a flow diagram of a method for determining user demand in accordance with an embodiment of the present invention.
FIG. 2 is a flowchart for generating a user search association rule as a method for determining a user demand according to an embodiment of the present invention.
3 is an illustration of an apparatus for determining user demand according to an embodiment of the present invention.
In the accompanying drawings, the same or similar reference numerals indicate the same or similar members.

It is to be noted that prior to proceeding with a more detailed description of the exemplary embodiment, some embodiments are described in terms of a process or method illustrated in the flowchart. Although the flowcharts have described each operation as a sequential process, considerable operations therein may be performed in parallel, in parallel, or concurrently. Here, the order of each operation can be rearranged. When these operations are completed, the processing can be stopped, but there may be additional steps not included in the figure. The processing may correspond to a method, a function, a procedure, a subroutine, a subprogram, and the like.

Referred to herein as a " computer device " refers to an intelligent electronic device that is capable of executing certain processes, such as numerical computing and / or logic computing, by executing certain programs or instructions, And may execute a predetermined processing process by executing a surviving instruction stored in advance in the storage device by the processor or may execute a predetermined processing process with hardware such as ASIC, FPGA, DSP, or the like, Lt; / RTI > Computer devices include, but are not limited to, servers, personal computers, laptops, tablet PCs, smart phones, and the like.

The computer device includes a user device and a network device. Here, the user equipment includes a computer, a smart phone, a PDA, and the like. The network equipment may be a single network server, a server group composed of a plurality of network servers, or a cloud computing-based large- But not limited to, a cloud, where cloud computing is one of distributed computing, a super virtual machine consisting of a group of computers in a loose coupling. Here, the computer device may be operated independently to realize the present invention, and the present invention may be realized through an interaction operation with other computer devices on the network by connecting to a network. Here, the network in which the computer device is located includes, but is not limited to, the Internet, WAN, MAN, LAN, VPN network, and the like.

Specifically, the user equipment, the network equipment, and the network are merely examples, and any other existing or future computer equipment or network should be used within the protection scope of the present invention, if applicable to the present invention, Which is hereby incorporated by reference.

The method described here (some of which are shown in the flowchart) may be implemented in hardware, software, firmware, middleware, microcode, hardware description language, or any combination thereof. When implemented in software, firmware, middleware, microcode, program code or code segments for executing the required tasks may be stored in a machine or computer readable medium (e.g., a storage medium). (One or more) processors may execute mandatory tasks.

The details of the specific structure and functionality disclosed herein are merely exemplary and are intended to illustrate exemplary embodiments of the invention. However, the present invention can be embodied in various alternative forms, and is not to be construed as being limited to the embodiments described herein.

Although the units have been described using the terms " first ", " second ", etc., it should be understood that these units are not limited to such terms. These terms are used only to distinguish one unit from another. For example, a first unit may be referred to as a second unit, and a second unit may be referred to as a first unit, as long as the scope of the exemplary embodiment does not depart from the scope. The term " and / or " as used herein includes any or all combinations of one or more of the listed related items.

When one unit is referred to as being " connected " or " coupled " to another unit, it may be directly coupled or coupled to the other unit, or an intermediate unit may be present. Correspondingly, when one unit is referred to as being " directly connected " or " directly coupled " to another unit, there is no intermediate unit. Other words to describe the relationship between units (eg, "located between ..." and "directly between ...", "adjacent to ..." and "directly adjacent to ..." .

The terminology used herein is for the purpose of describing specific embodiments only, and is not intended to limit the exemplary embodiments. Unless specifically stated otherwise in the context, the singular forms "a", "an" and "the" The use of the terms " comprises " and / or " comprising " as used herein should be interpreted as defining the presence of stated features, integers, steps, operations, units and / But does not preclude the presence or addition of other features, operations, units, members and / or combinations thereof.

It should be noted that, in some alternative implementations, the described functions / operations may occur in an order different from that shown in the figures. For example, depending on the function / operation involved, the two figures shown in sequence may in fact be executed substantially concurrently or in reverse order.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

1 is a flow chart of a method for determining a user demand according to an embodiment of the present invention,

S101: Querying a user search association rule based on a corresponding time sequence in accordance with a past search history of a user; And

S102: Determining the demand information of the user according to the user search association rule.

Compared with the prior art, the method according to the present invention has the following advantages. The time series is a sequence formed by arranging the numerical values of the same statistical index according to the temporal order of occurrence of the corresponding occurrence, and the time series based analysis is used to determine the user's demand using the time sequence (or dynamic sequence) The main objective is to make forecasts for the future based on existing historical data; Specifically, when people actually search for goods and services, the demand for information often has a temporal order in time, so that the method according to the present invention is based on time sequences, and when a user browses a web page, According to the past search history of the user and association rules set in advance, information related to some past search records can be pushed forward, so that users can find information they need more conveniently and quickly.

FIG. 2 is a flowchart for generating a user search association rule as a method for determining a user demand according to an embodiment of the present invention. 1 and Fig. 2 will be described below to explain the embodiments.

Here, before the step S101 of querying the association rule based on the corresponding time sequence according to the past record of the user shown in FIG. 1,

S201: acquiring the past search time and keyword of the user from the past record of the user when the inquiry request is received from the user;

S202: generating a time sequence of the past search of the user including the keywords arranged according to the past search time according to the past search time and keyword of the user;

S203: measuring the similarity of the time sequence; And

S204: generating association rule of user search according to the measurement result.

Here, there are many kinds of past records, and even if there is no back and forth order between past records, there is necessarily a causal relationship between past records. The present technical idea is to measure the similarity of time sequences, Causal relationships, and time sequences in which temporal precedence sequences exist together to improve the accuracy of pushing demand information in the future.

More specifically, the step S204 of generating a user search association rule according to the measurement result includes:

Generating a set of pattern subsequences represented as a pattern based on the time sequence;

Calculating a similarity between pattern subsections of the set; And

And generating a user search association rule based on the similarity between the pattern subsequences.

In more detail, in the method according to the present invention, a degree of similarity measurement function is defined using an algorithm such as a monotonic distance method or a vector distance algorithm to calculate a degree of similarity between pattern subsequences, Here, the algorithm is not limited to the forged distance method and the vector distance method described above. After measuring the similarity, it is determined which pattern subsequences can be integrated into the same frequent pattern, a frequent pattern is acquired using a method such as Apriori algorithm, and a corresponding association rule is generated, Is not limited to the Apriori algorithm described above. Although the time sequence refers to a sequence formed by arranging the numerical values of the same statistical index in accordance with the order of occurrence before and after the occurrence of the sequence, the similarity measure can not be directly conducted in many cases or the similarity measurement is relatively difficult. You can easily calculate the similarity by converting it to a pattern subsequence. The Apriori algorithm is an algorithm that mines a frequent item set of association rules. It is a core concept to mine frequent item sets through two steps: creation of candidate sets and downward closed detection of the plots. do.

Herein, the sequence pattern indicates that a set consisting of different sequences is given, wherein each sequence is arranged in order of different elements, each element (transaction) is composed of different items, and simultaneously One minimum supported threshold value specified by the user is given. Sequence pattern mining is to find all frequent subsequences, that is, the subsequence finds all frequent subsequences whose frequency in the sequence set is not lower than the user-specified minimum support threshold. The present invention refers to such a sequence pattern.

Alternatively, as shown in FIG. 2, after step S204 of creating a user search association rule,

S205: Proceeding with the generated user search association rule; And

S206: Selecting partial or all user search association rules from the generated user search association rule based on the verification result. Concretely, in step S206, test data verification and artificial evaluation are associated with each other, evaluation and verification are performed on the generated user search association rule, and by storing a rule having a relatively high interpretability and reliability, And can help the user to quickly determine demand information.

Alternatively, in the step of selecting the partial or all of the user retrieval association rules, the validity of the user retrieval association rule is detected according to a preset time interval, the removed user retrieval association rule is removed, Or all user search association rules can be suspended.

Specifically, the operation of detecting the validity of the user search association rule includes the following steps.

When the change in the information CTR is monitored and the change in the CTR exceeds a predetermined threshold value, it is determined that the corresponding association rule has been invalidated. If the information CTR has a change exceeding a predetermined threshold value, the current or corresponding association rule is already invalidated or the future orientation is insufficient, thereby easily adding and updating the association rule by removing the invalid association rule In addition, a better and faster service is provided to the user with an association rule having explanatory feasibility and high reliability, so that the user can conveniently use the service. We use ADWIN sliding window algorithm to monitor the change of information CTR.

Optionally, the step of measuring the similarity of the time-

Pre-processing the time sequence and removing noise of the data to generate a noise-removed time sequence; And

And measuring the similarity of the noise-removed time sequence.

Here, since the time sequence includes not only the past record contents of the user as we need but also inevitably contains interference data other than some valid data, the noise elimination processing proceeding with respect to the time sequence is performed based on the subsequent time sequence based processing And operations can be efficiently performed, and can also help to improve the accuracy of determining user demand information.

Alternatively, the removed noise is noise data that is out of the expected value calculated by a likelihood ratio or probability statistics method. Through the calculation, we can find data that is out of some expectation, and this data is noise data that should be removed soon.

When the degree of the data out of the expectation reaches a predetermined threshold value, the corresponding data can be regarded as noise data. Some data may be out of expectations, but may not exceed expectations too much. If this is the case, the data is likely to be valid data and is simply out of the expectation due to errors in the transmission and calculation process. Thereby helping to hold some valid data in which an error exists and reduce the situation in which valid data is removed resulting in greater errors in subsequent processing and computation.

Alternatively, the step of preprocessing the time sequence and removing the noise of the data to generate the noise-removed time sequence, the compression process is first performed on the noise-removed time sequence to generate a compressed time sequence and;

The similarity of the compressed time sequence can then be measured.

Here, only time sequences, especially a large number of time sequences, can ensure the accuracy of information orientation and display, but large amounts of data can cause problems such as data congestion, processor overload and degradation of processing speed; By carrying out the compression process on the time sequence, the processing time in association rule mining can be shortened; In the compression process, the time sequence can be converted into a subsequence represented by a series of patterns, which is helpful for subsequent rule mining.

Alternatively, the compression process uses an average compression method or a segment compression method.

Optionally, the user demand information determined according to the user search association rule comprises user demand information in at least two different time periods of the future; Here, after the above-described step S102,

And providing corresponding user demand information within a different time period to a client using the user and displaying the same.

Specifically, the user demand information determined in accordance with the user search association rule includes user demand information in at least two different time periods of the future, and automatically loads and displays each of them in different time periods. For example, if someone searches the Internet for bumper stickers, car seats, and other items on the Internet, it often means that he has just bought a new car. If this is the case, within the next month, he may have to purchase more goods for the new car, such as travel recorders, tire pressure gauges and fire extinguishers. In the next few months, as the new car enters the maintenance period, there is a possibility that it may be necessary to purchase items such as machine oil and machine filters. By automatically loading and matching different demand information according to different time periods, it is possible to allow the user to view information more in accordance with his demand, thereby making it convenient for the user, and at the same time, Make it visible to the crowd needed. For advertisers, advertisements for effects are given preference for short-term demand, and display advertisements are preferred for medium and long-term demand. Accordingly, the present technical idea automatically loads and displays different demand information to the user according to the time period, thereby making it convenient for the user to find the demand information.

Optionally, the method further comprises:

Performing reliability measurement on demand information;

Determining whether a client used by a user is in a standby state when the reliability of the demand information reaches a preset condition; And

And providing the user demand information to the client and displaying the user demand information if the client is in a waiting queue.

Specifically, the waiting queue may be a PC or a mobile-side large data transmission window, a web page refresh window, and an application installation waiting window.

When the demand information is displayed on a web page, it can be displayed in the following manner.

The display progresses alternately at predetermined time intervals according to the time order of the demand of the user. For example, when displaying an advertisement in an alternating manner, the advertisement that satisfies the user's latest demand is displayed as a top priority.

Since the capacity of the web page is limited, it is possible to display more contents on a limited layout by alternately displaying in accordance with a predetermined time interval, and also it is possible to reduce the area required for displaying contents, You can avoid situations that are annoying by showing only the contents.

Optionally, when a plurality of pieces of demand information are displayed on a web page, they can be arranged from left to right or from top to bottom according to time order. For example, you can arrange your ads from left to right to show the most recent demand on the left-most. In addition, an alternative time of user demand may be indicated in the lower part of the advertisement on the time axis.

Optionally, the demand time information may be additionally indicated in displaying the advertisements.

Compared to a relatively dry rolling atmosphere, information that meets the user's future demand can be automatically opened and displayed using these instruments. For example, you can display animated and sporty ads (mini-games). This method can enhance the impression of the user, reduce the user's irritation, improve the user experience, and improve the conversion rate.

FIG. 3 is a device for determining a user demand according to another embodiment of the present invention,

An apparatus (abbreviated as "inquiry apparatus 2") configured to inquire a user search association rule based on a corresponding time sequence according to a user's past search history; And

And a device (abbreviated as "demand information confirmation device 3") configured to determine the demand information of the user according to the user search association rule.

Compared with existing technologies, the device according to the present invention has the following advantages. The time sequence is a sequence formed by arranging the numerical values of the same statistical index according to the temporal order of occurrence, and the time sequence based analysis is based on the existing sequence index The main objective is to make forecasts for the future based on historical data of the future; Specifically, when people actually search for goods and services, the demand for information often has an order in time before and after. Therefore, the device according to the present invention is based on time sequence, and when a user browses a web page, According to the search history and association rules set in advance, information related to some past search records can be pushed forward, so that the user can find the information he / she needs more conveniently and quickly. The device according to the present invention generally uses a network device on the cloud side.

Optionally, the device 1 may,

An apparatus (not shown) configured to acquire a past search time and a keyword of the user from a past record of the user when receiving an inquiry request from the user;

A device (not shown) configured to generate a time sequence of past searches of the user including the keywords arranged according to the past search time according to the past search time and keyword of the user;

An apparatus (not shown) configured to measure the similarity of the time sequence; And

And an apparatus (not shown) configured to generate an association rule of a user search according to the measurement result.

Here, there are many kinds of past records, and causal relationships exist between past records even if there is no precedence sequence between past records. By measuring the similarity of the device time sequence according to the present invention, And the temporal order in which temporal precedence sequences exist, thereby improving the accuracy of pushing demand information in the future.

Here, the apparatus configured to generate the user search association rule according to the measurement result,

An apparatus (not shown) configured to generate a set of pattern subsequences represented as a pattern based on a time sequence;

An apparatus (not shown) configured to calculate a degree of similarity between pattern subsections of the set; And

And an apparatus (not shown) configured to generate a user search association rule based on the similarity between the pattern subsequences.

In more detail, in the method according to the present invention, the degree of similarity between pattern subsequences is calculated by defining a similarity measure function using an algorithm such as a forgone distance method or a vector distance method, Not limited to the street method; After determining the similarity measure, it is determined which pattern subsequences can be integrated into the same frequent pattern, and a frequent pattern is obtained using a method such as Apriori algorithm and a corresponding association rule is generated, . Although the time series indicates the sequence formed by arranging the numerical values of the statistical indices having the same time sequence in accordance with the temporal order of occurrence of the corresponding occurrence, the time sequence can not directly measure the similarity in many cases, or the similarity measurement is relatively difficult, First, by converting it to a pattern subsequence, the similarity can be conveniently calculated. The Apriori algorithm is an algorithm for mining a frequent item set of association rules. It is based on the concept of mining a frequent item set through two steps: generation of a candidate set and detection of a downward closure of a plot.

Optionally, the apparatus configured to generate a user search association rule according to the measurement result,

And an apparatus (not shown) configured to conduct verification for the generated user search association rule,

Here, the apparatus configured to perform the verification for the generated user search association rule selects partial or all user search association rules from the generated user search association rule based on the verification result.

Specifically, the apparatus configured to perform the verification for the generated user search association rule includes a user search association rule generated by associating test data verification and artificial evaluation so as to guide the information orientation and help the user to quickly determine demand information, , And can store a rule having relatively high interpretability and reliability among them.

Alternatively, the apparatus configured to perform the verification for the generated user search association rule may further be configured to detect the validity of the user search association rule according to a predetermined time interval, and to remove the user search association rule that has been invalidated.

Specifically, the operation procedure for checking the validity of the user search association rule is as follows.

When the change in the information CTR is monitored and the change in the CTR exceeds a predetermined threshold value, it is determined that the corresponding association rule has been invalidated. If the information CTR has a change exceeding a predetermined threshold value, the current or corresponding association rule is already invalidated or the future orientation is insufficient, thereby easily adding and updating the association rule by removing the invalid association rule In addition, a better and faster service is provided to the user with an association rule having explanatory feasibility and high reliability, so that the user can conveniently use the service. We use ADWIN sliding window algorithm to monitor the change of information CTR.

Optionally, the apparatus configured to measure the similarity of the time sequence,

A sub-device (not shown) configured to pre-process the time sequence and remove the noise of the data to generate a noise-removed time sequence; And

And a sub-device (not shown) configured to measure the degree of similarity of the noise-removed time sequence.

Here, since the time sequence includes not only the past record contents of the user as we need but also inevitably contains interference data other than some valid data, the noise elimination processing proceeding with respect to the time sequence is performed based on the subsequent time sequence based processing And operations can be efficiently performed, and can also help to improve the accuracy of determining user demand information.

Alternatively, the removed noise is noise data that is out of the expected value calculated by a likelihood ratio or probability statistics. Through the calculation, we can find data that is out of some expectation, and this data is noise data that should be removed soon.

When the degree of the data out of the expectation reaches a predetermined threshold value, the corresponding data can be regarded as noise data. Some data may be out of expectations, but may not exceed expectations too much. If this is the case, the data is likely to be valid data and is simply out of the expectation due to errors in the transmission and calculation process. Thereby helping to hold some valid data in which an error exists and reduce the situation in which valid data is removed resulting in greater errors in subsequent processing and computation.

Optionally, the apparatus configured to measure the similarity of the time sequence comprises:

A sub-device (not shown) configured to pre-process the time sequence to remove noise of the data;

A sub-device (not shown) configured to generate a noise-removed time sequence;

A sub-device (not shown) configured to perform a compression process on the time-sequence from which the noise is removed to generate a compressed time sequence; And

And a sub-device (not shown) configured to measure the similarity of the compressed time sequence.

Here, only time sequences, especially a large number of time sequences, can ensure the accuracy of information orientation and display, but large amounts of data can cause problems such as data congestion, processor overload and degradation of processing speed; By carrying out the compression process on the time sequence, the processing time in association rule mining can be shortened; In the compression process, the time sequence can be converted into a subsequence represented by a series of patterns, which is helpful for subsequent rule mining.

Alternatively, the compression process uses an average compression method or a segment compression method.

Optionally, a device configured to determine user demand,

(Abbreviated as "display device 4") configured to provide and display corresponding user demand information in a different time period to a client used by the user, respectively.

Specifically, the apparatus configured to determine the user demand information according to the user search association rule determines user demand information within two different time periods in the future. For example, if someone searches the Internet for bumper stickers, car seats, and other items on the Internet, it often means that he has just bought a new car. If this is the case, within the next month, he may have to purchase more goods for the new car, such as travel recorders, tire pressure gauges and fire extinguishers. In the next few months, as the new car enters the maintenance period, there is a possibility that it may be necessary to purchase items such as machine oil and machine filters. By automatically loading and matching different demand information according to different time periods, it is possible to allow the user to view information more in accordance with his demand, thereby making it convenient for the user, and at the same time, Make it visible to the crowd needed. For advertisers, advertisements for effects are given preference for short-term demand, and display advertisements are preferred for medium and long-term demand. Accordingly, the present technical idea automatically loads and displays different demand information to the user according to the time period, thereby making it convenient for the user to find the demand information.

Optionally, the apparatus for determining user demand comprises:

An apparatus configured to perform reliability measurement on demand information; Further comprising:

When the reliability of demand information reaches a preset condition, it is judged whether or not the client used by the user is in a standby state;

If the client is in the waiting queue, the user demand information is provided to the client and displayed.

Specifically, the waiting queue may be a PC or a mobile-side large data transmission window, a web page refresh window, and an application installation waiting window.

When the demand information is displayed on a web page, it can be displayed in the following manner.

The display progresses alternately at predetermined time intervals according to the time order of the demand of the user. For example, when displaying an advertisement in an alternating manner, the advertisement that satisfies the user's latest demand is displayed as a top priority.

Since the capacity of the web page is limited, it is possible to display more contents on a limited layout by alternately displaying in accordance with a predetermined time interval, and also it is possible to reduce the area required for displaying contents, You can avoid situations that are annoying by showing only the contents.

Optionally, when a plurality of pieces of demand information are displayed on a web page, they can be arranged from left to right or from top to bottom according to time order. For example, you can arrange your ads from left to right to show the most recent demand on the left-most. In addition, an alternative time of user demand may be indicated in the lower part of the advertisement on the time axis.

Optionally, the demand time information may be additionally indicated in displaying the advertisements.

Compared to a relatively dry rolling atmosphere, information that meets the user's future demand can be automatically opened and displayed using these instruments. For example, advertisements (mini games) can be displayed in a lively and sporty format. This method can enhance the impression of the user, reduce the user's irritation, improve the user experience, and improve the conversion rate.

It should be noted that the present invention may be implemented in software and / or a combination of software and hardware and may be implemented using, for example, an application specific integrated circuit (ASIC), a general purpose computer, or any other similar hardware device. In one embodiment, a software program of the present invention may be executed by a processor to perform the steps or functions described above. Likewise, a software program (associated data structure) of the present invention may be stored on a computer-readable recording medium and stored on, for example, a RAM storage device, a magnetic or optical drive device, or a floppy diskette and similar device. In addition, some of the steps or functions of the present invention may be implemented using hardware, for example, a circuit that executes each step or function in combination with a processor.

It will be understood by those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is self-evident. Accordingly, in all respects, the embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention should not be limited by the above description but should be defined in the appended claims, All changes that come within the meaning and range of the requirements are to be embraced within the invention. Any reference signs in the claims shall not be construed as limiting the claim. It is also to be understood that the term " comprises " does not exclude other units or steps, and that an odd number does not exclude a plurality. The plurality of units or devices described in the device claims may be implemented in software or hardware as a unit or device. The terms " first ", " second ", and the like denote names but do not indicate any particular order.

Claims (19)

Querying a user search association rule based on a time sequence based on a user's past search history; And
And determining the demand information of the user according to the user search association rule. The method according to claim 1,
Before the step of querying the association rule based on the corresponding time sequence according to the past record of the user,
Acquiring the past search time and keyword of the user from the past record of the user when an inquiry request is received from the user;
Generating a time sequence of the user's past search including a keyword arranged according to a past search time according to the past search time and keyword of the user;
Measuring the similarity of the time sequence; And
And generating an association rule of a user search according to the measurement result. 3. The method of claim 2,
Wherein the generating of the association rule of the user search according to the measurement result comprises:
Generating a set of pattern subsequences represented as a pattern based on the time sequence;
Calculating a similarity between the pattern subsequences of the set; And
And generating a user search association rule based on the similarity between the pattern subsequences. The method according to claim 2 or 3,
Wherein the generating of the association rule of the user search according to the measurement result comprises:
Proceeding with the verification of the generated user search association rule; And
And selecting partial or all user search association rules from the generated user search association rule based on the verification result. 5. The method according to any one of claims 2 to 4,
Wherein the step of measuring the similarity of the time sequence comprises:
Pre-processing the time sequence and removing noise of the data to generate a noise-removed time sequence; And
And measuring the similarity of the noise-removed time sequence. ≪ Desc / Clms Page number 19 > 6. The method according to any one of claims 2 to 5,
Wherein the step of measuring the similarity of the time sequence comprises:
Pre-processing the time sequence and removing noise of the data to generate a noise-removed time sequence;
Performing compression processing on the noise-removed time sequence to generate a compressed time sequence; And
And measuring the similarity of the compressed time sequence. ≪ RTI ID = 0.0 > 11. < / RTI > 7. The method according to any one of claims 1 to 6,
Wherein the user demand information determined according to the user search association rule includes user demand information in at least two different time periods in the future,
And providing the corresponding user demand information to the client using the user in different time periods and displaying the user demand information. 8. The method according to any one of claims 1 to 7,
Performing reliability measurement on the demand information;
Determining whether a client used by a user is in a standby state when the reliability of the demand information reaches a preset condition; And
And providing and displaying the user demand information to the client when the client is in the waiting queue. An apparatus configured to query a user search association rule based on a time sequence based on a past search history of a user; And
And a device configured to determine demand information of the user according to a user search association rule. 10. The method of claim 9,
An apparatus configured to acquire the user's past search time and keyword from the past record of the user when receiving a query request from the user;
An apparatus configured to generate a time sequence of the user's past search including a keyword arranged according to a past search time according to the past search time and keyword of the user;
An apparatus configured to measure a similarity of the time sequence; And
And a device configured to generate an association rule of a user search according to the measurement result. 11. The method of claim 10,
An apparatus configured to generate an association rule of a user search according to the measurement result,
An apparatus configured to generate a set of pattern subsequences represented as a pattern based on a time sequence;
An apparatus configured to calculate a degree of similarity between pattern subsequences of the set; And
And a device configured to generate a user search association rule based on the similarity between the pattern subsequences. The method according to claim 10 or 11,
An apparatus configured to generate an association rule of a user search according to the measurement result,
A device configured to proceed with verification for the generated user search association rule; And
And a device configured to perform the verification on the generated user search association rule and to select partial or all user search association rules from the generated user search association rule based on the verification result . 13. The method according to any one of claims 10 to 12,
An apparatus configured to measure the similarity of the time sequence,
A sub-device configured to process the time sequence and generate noise-canceled time sequence by removing noise of the data; And
And a sub-device configured to measure the similarity of the noise-removed time sequence. 14. The method according to any one of claims 10 to 13,
An apparatus configured to measure the similarity of the time sequence,
A sub-apparatus configured to pre-process the time sequence to remove noise of the data;
A sub-apparatus configured to generate a noise-removed time sequence;
A subdevice configured to perform a compression process on the noise-removed time sequence to generate a compressed time sequence; And
And a sub-device configured to measure the similarity of the compressed time sequence. 15. The method according to any one of claims 9 to 14,
Wherein the user demand information determined according to the user search association rule includes user demand information in at least two different time periods in the future,
Further comprising a device configured to provide and display corresponding user demand information to the user using the user within a different time period. ≪ RTI ID = 0.0 > 31. < / RTI > 16. The method according to any one of claims 9 to 15,
An apparatus configured to perform reliability measurement on the demand information;
An apparatus configured to determine whether a client used by a user is in a standby state when the reliability of the demand information reaches a predetermined condition; And
Further comprising: a device configured to provide and display the user demand information to the client when the client is in a waiting queue. In a non-volatile computer readable medium,
Wherein the non-volatile computer-readable medium comprises computer code, wherein when the computer code is executed, the method of any one of claims 1 to 8 is executed. In a computer program product,
9. A computer program product, characterized in that the method of any one of claims 1 to 8 is executed when the computer program product is executed by a computer device. A computer device comprising:
A storage device; And
A processor,
The computer code is stored in the storage device,
Wherein the processor is configured to execute the computer code and to execute the method of any one of claims 1-8.

Images