KR20050102869A - Method and system for detecting serach terms whose popularity increase rapidly - Google Patents

Abstract

The present invention relates to a method and system for detecting a rapidly rising search word from a user terminal, and more particularly, to generate a rising standard value by using an average value and a standard deviation of each search word, and to measure the rising standard value and the measurement. The present invention relates to a method and a system capable of detecting a search word in which the number of inputs surges by comparing the number of inputs.

According to an aspect of the present invention, there is provided a method of detecting a rising search word, the method including maintaining a database including a search word identifier and first input count information for a first search word associated with the search word identifier (the first input count information is generated at predetermined time intervals). Recorded in the database), generating average value information and standard deviation information regarding the first input frequency information for each of the first search terms by referring to the database; and using the average value information and the standard deviation information, the first search word. Generating rapidly reference numerical information for each step, receiving a second search term from a user terminal, generating second input count information by counting the number of inputs per time interval for each of the second search terms, and receiving the second number of inputs. Comparing the information with the sudden reference numerical information, and the And detecting a third search word associated with third input count information that is equal to or higher than the sudden reference number information among second input count information.

According to the present invention, there is provided a method and system for detecting a rising search word that can quickly detect a search word in which the number of inputs from the user rises rapidly and provide a customized service accordingly.

Description

Rising search term detection method and system {METHOD AND SYSTEM FOR DETECTING SERACH TERMS WHOSE POPULARITY INCREASE RAPIDLY}

The present invention relates to a method and system for detecting a rapidly rising search word from a user terminal, and more particularly, to generate a rising standard value by using an average value and a standard deviation of each search word, and to measure the rising standard value and the measurement. The present invention relates to a method and a system capable of detecting a search word in which the number of inputs surges by comparing the number of inputs.

In general, a search service system that provides a search service includes a search result corresponding to the search word when a user inputs a search word (for example, a web site including the search word, an article including the search word, and a file name including the search word). To the user).

In such a search service system, there may be various search terms input from a user, but in reality, the ratio of more than 1000 top search terms to the total search requests according to the number of inputs from the user is very high, and the rest of the search terms are found in the entire search request. The proportion is very low. This means that a higher level of search service can be provided to the user through a customization service that provides a search result intensively for some of the top search terms having a high number of search requests.

However, when analyzing the top search terms, 1) there are regular search terms (eg, "game", "go-stop", "celebrity", etc.), while 2) the ranking rises in a certain period. Sometimes it belongs to a top search term.

In the case of 1), most Internet search service providers currently provide optimized information to users. However, in case of 2), it is difficult to recognize and detect a search word that rapidly increases in the number of inputs from the user. As a result, Internet search service providers have not responded appropriately and often realized that after a considerable time. In this case, when the Internet search service provider recognizes the response and attempts to respond, there is a case in which the response becomes meaningless because the ranking of the corresponding search word has already decreased.

In addition, in the case of 2), a person may use a method of identifying a term appearing in a newspaper article or the like and estimating it as a rising search word. However, detection of the rising search word by such a method may be performed only after the search word is widely known to the public. As a result, there is a disadvantage in that the detection is very late in time, and there is also a problem in that it is impossible to detect all search terms in all fields.

On the other hand, Korean Patent Application No. 10-2000-7010220 (Publication No. 2001-0086259) is an Internet search engine that provides searchers with a high-flyers hit list of rapidly increasing popularity web pages. Is disclosed. The high flyer hit list refers to a list of sites in which the number of clicks by a user in a search result increases rapidly.

The Internet search engine measures the number of clicks for each site and the cumulative click counts for each site so far during the initial period, and calculates the rate of change of the popularity of each site, thereby rapidly increasing the popularity. Provides searchers with a high flyer hit list.

However, even when using the Internet search engine, there is still a problem that it is not possible to detect whether the number of inputs for a specific search word is soaring by detecting whether the popularity of the clicked site is rapidly rising, not the search word input from the user. do.

After all, even with the Internet search engine, the Internet search service provider is still unable to detect a search term belonging to a higher search term due to a sudden rise in the ranking in a certain period of time, and thus, after a considerable time has elapsed without appropriate response. Recognizing this, there is a disadvantage in that it is not possible to provide a user with a customization service that provides a search result intensively for a search word in which the number of search requests is rapidly increasing at an appropriate time.

Accordingly, there has been a demand for the emergence of a new technology capable of quickly detecting a search word in which the number of inputs from the user is rapidly rising, thereby providing a user with a customized service for the rising search word to maximize the user's convenience.

The present invention has been made to solve the problems of the prior art as described above, the number of inputs is increased by generating a jump reference value using the average value and standard deviation of each search word and comparing the jump reference value and the measured number of inputs It is an object of the present invention to provide a method and a system for detecting a rising search word that can provide a user with a customized service for the rising search word by detecting a rising search word.

In addition, the rapidly rising search word detection method and system according to the present invention by quickly detecting a search word that the number of inputs from the user rapidly provides to the user, the user's desire to quickly receive a search result using a rapidly rising search term To satisfy them.

In addition, the rapid search keyword detection method and system according to the present invention to quickly detect a search word that is rapidly rising the number of inputs from the user, the Internet search service provider to maximize the profit by utilizing the detected rising search term in various fields Its purpose is to make it possible.

In addition, the method and system for detecting a rising search word according to the present invention aggregates the data accumulated over all fields for a certain period of time to detect the rising search word, thereby satisfying the user's desire to receive a rising search word more accurately and in various fields. For that purpose.

In order to achieve the above object and to solve the above-described problem of the prior art, the method of detecting a rising search word according to an embodiment of the present invention includes a search word identifier and first input count information for a first search word associated with the search word identifier. Maintaining a database including the first number of times input information is generated at predetermined time intervals and recorded in the database; an average value of the first number of times input information and standard for each of the first search terms with reference to the database; Generating deviation reference information, generating rapid reference numerical information for each of the first search terms using the average value information and the standard deviation information, receiving a second search term from a user terminal, and the time for each second search term Generating second input count information by counting the number of inputs per interval; And comparing the second number of times of input information with the spike reference number information, and detecting a third search word associated with the third number of times of input information that is equal to or higher than the number of spikes reference value information among the second number of times of input information. It is characterized by.

According to one aspect of the invention, maintaining a second database including the search term input ranking information according to the first number of times of input information (the search term input ranking information is generated at each time interval is recorded in the second database), Calculating input frequency increase information associated with the third input frequency information with reference to the database; retrieving input rank information corresponding to the third input frequency information with reference to the second database; and the second database. Calculating input rank increase information associated with the input rank information with reference to generating standard normal variable information about the third input count information using the average value information and the standard deviation information, and increasing the number of input counts Information, the input ranking information, the input ranking rising information, and the table Generating spike index information using regular variable information, and generating spike rank information on the third search word based on the spike index information. .

In addition, the rapidly rising search word detection system according to the present invention includes a database including a search word identifier and the first number of times input information for the first search term associated with the search term identifier, the database by generating the first number of times input information at a predetermined time interval Means for generating average value and standard deviation information for generating average value information and standard deviation information about the first input frequency information for each of the first search terms by referring to the database; and the average value information and the standard deviation information. A quick reference numerical information generating means for generating sharp reference numerical information for each of the first search terms, a search term input means for receiving a second search term from a user terminal, and counting the number of inputs per time interval for each of the second search terms; 2 to generate input count information Search word detection for detecting a third search word associated with a counter means, comparison means for comparing the second input count information with the jump reference numerical information, and third input count information that is equal to or higher than the jump reference numerical information among the second input count information; It comprises a means.

Hereinafter, a method and system for detecting a rising keyword according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram illustrating a network connection of a fast rising keyword detection system according to an exemplary embodiment of the present invention. Users access the wired / wireless communication network to the rising search word detection system 100 using the user terminal 110a or 110b and input a search word. The rising search word detection system 100 records the number of inputs for the search word in a predetermined database and detects the rising search word by analyzing the search word. In addition, the rising search word detection system 100 according to another embodiment of the present invention may generate rank information on the detected rising search word and transmit it to the user terminal 110a or 110b. The rising search word detection system 100 according to the present invention may be integrated and operated in a search service system, and thus, when a user accesses an Internet search service system, ranking information regarding the detected rising search word is provided to the user. can do. In addition, when the search word input from the user corresponds to the detected rising search word, a so-called personalized service may be provided to the user by providing detailed search results prepared in advance.

2 is a flowchart illustrating a method of detecting a rising keyword according to an exemplary embodiment of the present invention. The rising keyword search method according to the present embodiment is provided by a predetermined rising keyword detection system.

In step 201, the rapidly rising search word detection system according to the present invention maintains a database including a search word identifier and first input count information for a first search word associated with the search word identifier. The rapidly rising search word detection system generates the first input count information at predetermined time intervals and records the first input number information in the database.

3 is a diagram illustrating an example of the database according to an embodiment of the present invention. As illustrated in FIG. 3, the database may include a search term identifier 301 and information on the number of times of input of a search term associated with the search term identifier 301. Referring to reference numeral 302 shown in FIG. 3, the database includes input count information "50", "60", and "500" for the search term identifier "s54095" and "snowboard" associated with "snowboard". It is recorded every predetermined time interval. Although FIG. 3 illustrates an example of a database in which the time interval is "day," the time interval may be variously determined by the system operator, such as "day", "two days", "weekly", and the like. You can also change the time interval to different time intervals.

In step 202, the rapidly rising search word detection system generates the average value information and the standard deviation information of the first input frequency information for each of the first search words by referring to the database. According to an embodiment of the present invention, the sudden rising search word detection system may generate the average value information and the standard deviation information at predetermined time intervals, wherein the time interval is “1 month”, “2” by the system operator. Month, "" three months, "and the like, and the system operator may change the existing time interval to another time interval. Also, the sudden rising search word detection system may generate average value information and standard deviation information regarding the first input frequency information at an arbitrary time point.

According to an embodiment of the present invention, in generating the average value information and the standard deviation information for each keyword in the step 202, the spike search term detection system may not consider data at the time of determining whether the spike keyword is a rising keyword. . That is, as will be described later, the rising search word detection system receives a second search word from a user terminal and determines whether the second search word is a rising search word. The data regarding the number of inputs of the second search word includes average value information and standard deviation. It may not be taken into account in generating the information. Since the number of inputs at the time corresponding to the rapidly rising search word will be significantly larger than the previous data, in consideration of this, since the average value information and the standard deviation information regarding the number of inputs of the second search word become very large, the second search word at the present time This is because it is not possible to accurately determine whether the number of inputs to is rapidly increasing. Therefore, according to the present embodiment, it is possible to obtain the effect of enabling more accurate detection of the rising search word.

In step 203, the rising search word detection system generates rising reference numerical information for each of the first search words by using the average value information and the standard deviation information. The sudden reference value information is a reference value for determining whether the number of inputs for a specific search word is rapidly increasing.

According to an embodiment of the present invention, the spike search word detection system may generate the spike reference numerical information by adding the average value information and the value obtained by multiplying the standard deviation information by a predetermined number. For example, the predetermined number may be three. In this case, the sudden reference numerical information may be expressed by a formula such as "mean + (3 * standard deviation)".

According to this embodiment, the number of inputs corresponds to more than the above-mentioned spike standard numerical information only when a search word having a significantly higher number of inputs than the number of inputs on average appears, thereby preventing excessively rising search terms from being detected. As a result, an unnecessary memory requirement due to an excessive number of spiked search words is detected, and an improvement in processing speed of the spiked search word detection system can be achieved.

In step 204, the escalating search word detection system receives a second search word from a user terminal, and in step 205, counts the number of inputs per time interval for each second search word to generate second input count information. In step 206, the second number of times of input information is compared with the sudden increase standard value information.

In addition, the rising search word detection system detects a third search word in step 207 associated with the third input count information that is equal to or higher than the jump reference numerical information among the second input count information.

As performed in steps 204 to 207, the rapidly rising search word detection system compares a reference value based on previous data for each search word with the number of inputs at the current time for each search word, thereby As the number of times of input increases rapidly, a search word that records more than the reference value is detected as a rising search word.

4 is a diagram illustrating an example of a graph for detecting a rising search word through comparison with the rising reference numerical information as described above. Referring to FIG. 4, sudden reference numerical information for a specific search term is illustrated in a graph corresponding to reference numeral 401, and input count information for the search term is illustrated in a graph corresponding to reference numeral 402. . At the time corresponding to the reference numeral 403, the number of times of input information for the search word is higher than the rising reference numerical information, and the rising search word detection system detects the search word as the rising search word at the time point.

According to an embodiment of the present invention, the spike search word detection system detects the third search word only when the third input count information is greater than or equal to a predetermined value in step 207. Is provided. That is, even when the number of inputs for a specific search word exceeds a reference value, the rapidly rising search word detection system may not detect a search word whose number of inputs does not reach a predetermined value as a sudden search word. An important reason for detecting spiked search terms is to quickly identify spiked search terms and provide them with customized services, so for those queries with a very small number of search requests, providing such customized services does not mean much. The rising search word detection system according to the present exemplary embodiment may detect only a search word having a predetermined number of inputs or more as a rising search word. For example, when the time interval is one day, the rising search word detection system may detect only a search word having 100 or more inputs per day as the rising search word.

According to an embodiment of the present invention, there is provided a method of detecting a rising search word capable of generating rank information of a detected rising search word by the system of detecting a rising search word. The spike search term detection method according to the present embodiment may further include steps 208 to 214.

In step 208, the rising search word detection system maintains a second database including search word input ranking information according to the first input count information. The rapidly rising search word detection system generates the search word input ranking information for each time interval and records the search word input ranking information in the second database.

FIG. 5 illustrates an example of the second database including the search word input ranking information according to an embodiment of the present invention. As illustrated in FIG. 5, the second database may include search term input and search term input ranking information according to input count information of a search term associated with the search term identifier. Referring to the reference numeral 501 of FIG. 5, the second database includes search term input ranking information based on the number of times of input of the search term identifier "s54095" and "snowboard" associated with "snowboard". "800" and "100th" are recorded at predetermined time intervals. As in FIG. 3, FIG. 5 shows an example of a database in which the time interval is “day”, but the time interval may be variously determined by the system operator such as “day”, “two days”, “one week”, and the like. In addition, the system operator may change the existing time interval to another time interval.

In step 209, the rapidly rising search word detection system calculates input count increase information associated with the third input count information by referring to the database. Since the database records the number of times of input of the search word generated at predetermined time intervals as described with reference to FIG. 3, the number of times of input is calculated by calculating a difference between the number of times of input at the present time and the number of times at the previous time. Ascent information can be calculated. For example, when the time interval is one day, when the number of yesterday's inputs for a specific search word is 100 and the number of today's inputs is 400, the input number rising information may be calculated as 300.

In step 210, the rising search word detection system searches for input ranking information corresponding to the third input count information by referring to the second database, and in step 211, the input ranking information by referring to the second database. Compute input rank increase information associated with. Since the second database records search word input ranking information generated at predetermined time intervals as described with reference to FIG. 5, the second database may search for an input rank corresponding to the number of inputs at the present time. The input rank increase information may be calculated by calculating a difference between an input rank and an input rank at a previous point in time. For example, when the time interval is one day, today's input number is 400 for a specific search term, and today's input rank is 5000, and yesterday's input rank is 10000, the input ranking information is searched as 5000. The input rank increase information may be calculated as 5000.

In step 212, the rapidly rising search word detection system generates standard normal variable information about the third input frequency information using the average value information and the standard deviation information generated in step 202.

In general, it is very difficult and cumbersome to integrate a normal distribution function each time to find a probability in a range of values that a random variable will have. Therefore, in this case, the normal variable is converted into a standard normal variable. It is very convenient to create and then find the probability you want. Here, the standardized variable means that the difference obtained by subtracting the average value from individual random variable values is divided into standard deviations and converted into new random variables. The newly converted random variable is referred to as a standard normal variable. In operation 212, the sudden rising search word detection system may generate the standard normal variable information as described above using the average information and the standard deviation information generated in operation 202.

In step 213, the rising search word detection system generates the rising index information by using the input number of rising information, the input ranking information, the input ranking rising information, and the standard normal variable information.

According to an embodiment of the present invention, the escalating search word detection system may be configured to multiply the multiplication operation value of the input number rising information, the reciprocal of the input ranking information, the input ranking rising information, and the standard normal variable information in step 213. The spike index information may be generated using the multiplication operation value. According to the present embodiment, the multiplication operation of the input count increase information, the input rank rise information, and the standard regular variable information may be multiplied to measure the degree of sudden increase in the number of inputs for a specific search word. This will allow you to discover more important search terms.

Also, for example, if there is a "A" search term ranked 100th to 10th and a "B" search term ranked 100th to 5000th, in general, the "A" search term is more meaningful. Although it can be said that the recorded search word, there is a possibility that the degree of sudden increase of the search word "B" is measured more and ranks higher. Therefore, the sudden rising search word detection system according to the present exemplary embodiment can obtain an effect of preventing an unreasonable result from being generated by multiplying the inverse of the input ranking information.

In addition, according to an embodiment of the present invention, the sudden rising search word detection system in generating a multiplication operation value of the input number of rising information, the reciprocal of the input ranking information, the input ranking rising information, and the standard normal variable information The method may include generating order information associated with the standard normal variable information and generating the multiplication operation value using the order information. In this case, the order information may vary according to a predetermined criterion. According to this embodiment, by adjusting the weight of the standard normal variable information by using the order information, it is possible to reduce the possibility that various types of unreasonable results can be derived when generating the rising rank information.

According to another embodiment of the present invention, the escalating search word detection system generates an addition operation value of the input count increase information, the input rank increase information, and the standard regular variable information in step 213, and the addition operation. The multiplication operation value of the value and the reciprocal of the input rank information may be generated, and the spike index information may be generated using the multiplication operation value.

In addition, according to the present embodiment, the rising search word detection system generates first weight information associated with the input number rising information and generates second weight information associated with the input rank rising information in generating the addition operation value. And generating third weight information associated with the standard normal variable information, and generating the addition operation value using the first weight information, the second weight information, and the third weight information. It is available.

According to the present embodiment, the degree of spike in the number of inputs for a specific keyword can be measured, and thus the degree of spike is larger and a more important keyword can be found. In addition, the sudden rising search word detection system according to the present exemplary embodiment may obtain an effect of preventing an unreasonable result as described above by multiplying the inverse of the input ranking information.

According to an embodiment of the present invention, the sudden rising search word detection system may vary the first weight information, the second weight information, or the third weight information according to a predetermined criterion. According to the present embodiment, by adjusting the weight of each factor by using the weight information, it is possible to reduce the possibility of various types of unreasonable results due to one factor having too much influence on the spike index information. .

According to another exemplary embodiment of the present disclosure, the rapidly rising search word detection system may include the first weight information, the second weight information, and the third weight information, respectively, the input number rising information, the input rank rising information, and the Standard normal variable information can be normalized to a value between 0 and 1, inclusive. According to the present embodiment, by adjusting the values of the input number rising information, the input rank rising information, and the standard normal variable information within a certain range, any one factor may appear as having an excessive influence on the sudden index information. That can prevent irrational results.

In operation 214, the rising search word detection system may generate rising rank information regarding the third search word based on the rising index information. The system operator may use the generated jump ranking information to provide a user with a more specific personalized service for a search word with a greater degree of spike and more important. In addition, the system operator can satisfy the user's intellectual desire to wonder what search terms are recently increasing in popularity by providing the generated rising ranking information to the user.

In the above, an embodiment of detecting a rising search word and ranking the detected rising search words has been described above. However, an embodiment of detecting the rising search word using the factors used to determine the ranking may exist. An example is demonstrated. 6 is a flowchart illustrating a method of detecting a rising keyword according to the present embodiment.

In step 601, the rising search word detection system according to the present embodiment maintains a first database including a search word identifier and first input count information for a first search word associated with the search word identifier. The first input count information is generated at predetermined time intervals and recorded in the first database.

In step 602, the rising search word detection system maintains a second database including search word input ranking information according to the first input count information. The search term input ranking information is generated at each time interval and recorded in the second database.

In operation 603, the rapidly rising search word detection system receives a second search word from a user terminal, and in step 604, counts the number of inputs per time interval for each second search word to generate second input count information.

In step 605, the escalating search term detection system calculates the input count increase information associated with the second input count information with reference to the first database, and in step 606 the second input with reference to the second database. Input rank information corresponding to the retrieval information is retrieved, and in step 607, the input rank increase information associated with the input rank information is calculated with reference to the second database.

In step 608, the rapidly rising search word detection system generates the average value information and standard deviation information about the first input frequency information for each of the first search terms by referring to the first database, and in step 609, the average value information and The standard normal variable information about the second input frequency information is generated using the standard deviation information.

In step 610, the spike search word detection system generates spike index information by using the input number of rising information, the input ranking information, the input ranking rising information, and the standard normal variable information, and in step 611 The rising keyword may be detected based on the rising index information.

According to the embodiment as described above, it is possible to detect the rising search word more accurately using a variety of factors. In addition, since the ranking information can also be generated by using the index information, there is an advantage that does not need to separately create the ranking information.

According to an embodiment of the present invention, 1) when the input rank increase information for the search word is a predetermined value or more, 2) when the search word newly enters within a predetermined rank according to the input rank for the search word, or 3) the search word. There is provided a method of calculating the input rank rising information for a predetermined period and detecting the search word as a rising search word when the input ranking rising information is continuously a predetermined value or more during the period.

In the case of 1), the rising search word detection system determines whether the input rank rising information for the search word input from the user terminal is equal to or greater than a predetermined value, and when the determination result is equal to or greater than the numerical value, the search word is detected as the rising search word. can do. In this case, the numerical value may be changed according to the input rank of the search word in order to prevent the rapidly rising search word having a low input rank.

In case of 2), the escalating search word detection system determines whether the search word newly enters within a predetermined rank by using input ranking information on the search word input from the user terminal. Search terms can be detected as rising search terms.

In case of 3), the rapidly rising search word detection system calculates the input ranking rising information for the search word input from the user terminal at each time interval in which the number of times of input information is generated for a predetermined period, and the input ranking rising information is generated during the period. It may be continuously determined whether the predetermined value or more, and as a result of the determination, if the predetermined value is continuously or more during the period, the search term may be detected as a rising search term. In this case, the numerical value may be changed according to the input rank of the search word in order to prevent the rapidly rising search word having a low input rank.

According to the embodiment 1) or 2), a search word whose input rank rises sharply can be easily detected as a rising search word, thereby simplifying a process for detecting a rising search word, thereby reducing unnecessary memory requirements and The speed is improved. In addition, according to the embodiment of 3), a search word having a continuous ranking increase for a predetermined period of time can be detected as a rising search word, so that a search word that is continuously increasing in number instead of a search word that is temporarily increased in number of times is input. The effect can be obtained.

According to an embodiment of the present invention, a method of detecting a rapidly rising search word by generating average value information and upper management limit information for each search word using an Exponentially Weighted Moving Average (EWMA) chart and determining whether the average value information is moved Although provided, this embodiment will be described below.

In general, control charts are widely used for various purposes, but among them, EWMA charts are most commonly used because they can be easily generalized and small changes are easily detected. In the manufacturing process, it is very important to continuously monitor the quality of the product during the production process in order to produce the product quality to the standard. EWMA control chart is generally used to solve this problem.

According to this embodiment, by using the EWMA control chart to detect the rising search word, it is possible to monitor the trend of the number of times of input with respect to any one of the search terms.

The spike search term extraction method according to the present embodiment can be made by the following statistical problem. First of all, the average number of times , Dispersion A normal search distribution according to the present invention, the rapidly rising search word detection system according to the present embodiment determines whether the average for the search word moving daily. The determining that the average is shifted at a specific point in time by the escalating search term detection system means that the user's interest in the search word is different from before, and the escalating search term detecting system may detect the search term as the rising search term at the time point. have.

This problem is solved using the EWMA chart. Considering the following statistic at time t,

Equation 1. Statistics >

statistic Is the number of inputs measured at time t. Calculate the weighting factor of p and calculate the weighting factor of (1-p) for the statistics just before. Through this, the average shift can be considered by giving more weight to recent measurements. Equation 1 can be rewritten as follows.

Equation 2. Statistics >

Meanwhile, statistics The variance of can be calculated as

Equation 3. Statistics Dispersion of>

In addition, using the general control chart theory, the following management upper limit can be calculated.

<Equation 4. Upper management limit>

In Equation 4, Is You can use 1.128 as a factor to make an uncomfortable estimate, Is the average of the moving ranges. In other words, , Can be expressed as

According to the method of detecting the rising search word using the EWMA control chart according to the present embodiment, the rising search word detection system is calculated every time If the average exceeds the UCL, it may be determined that the average has moved, and thus the search term thus determined may be detected as a rising search term.

As an example of a method of detecting a rising search word in consideration of the current data and the previous 90 days, the rising search word detection system includes: Current data 2) Calculate the average and upper limit of the EWMA chart using the previous 90 days of data. of To convert from It is possible to determine whether or not the average shifts by checking whether or not the management upper limit is exceeded.

Hereinafter, a method of detecting a rising search word according to the present embodiment will be described based on the above theoretical configuration. 7 is a flowchart illustrating a method of detecting a rising keyword according to the present embodiment.

In step 701, the rising search term detection system according to the present embodiment maintains a database including a search term identifier and first input count information for a first search term associated with the search term identifier. The first input count information is generated at predetermined time intervals and recorded in the database.

In operation 702, the rapidly rising search word detection system generates the average value information and the upper limit management information about the first input frequency information for each first search word by using an EWMA (Exponentially Weighted Moving Average) chart. .

In operation 703, the escalating search word detection system receives a second search word from a user terminal, and in operation 704, counts an input count per time interval for each second search word to generate second input count information.

In step 705, the escalating search word detection system determines whether the average value information is moved based on the second input frequency information and the management upper limit information, and when it is determined that the average value information is moved, step 706. In FIG. 3, a rising search term may be detected by detecting a third search term associated with the moved average value information.

In addition, embodiments of the present invention include computer-readable media containing program instructions for performing various computer-implemented operations. The program recorded on the computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. The media may be those specially designed and constructed for the purposes of the present invention, or they may be of the kind well-known and available to those having skill in the computer software arts.

Hereinafter, a rising keyword search system according to another embodiment of the present invention will be described. 8 is a block diagram illustrating a system for detecting a rising keyword according to the present embodiment.

The rapidly rising search word detection system 800 according to the present embodiment includes a database 801, a database management means 802, an average value / standard deviation information generating means 803, a rising reference numerical information generating means 804, and a search word input means ( 805, counter means 806, comparison means 807, and search term detection means 808.

The database 801 includes a search term identifier and first input count information for a first search term associated with the search term identifier. 3 is a diagram illustrating an example of the database 801. Since FIG. 3 has already been described herein, a detailed description thereof will be omitted.

The database management means 802 generates the first input count information at predetermined time intervals and records the first input count information in the database 801.

The mean / standard deviation information generating means 803 refers to the database 801 to generate average value information and standard deviation information about the first input frequency information for each of the first search terms. According to an embodiment of the present invention, the mean value / standard deviation information generating means 803 may not consider the data at the time of determining whether it is a rapidly rising search word in generating average value information and standard deviation information for each search word. .

The quick reference numerical information generating means 804 generates the quick reference numerical information for each of the first search terms by using the average value information and the standard deviation information. The sudden reference value information is a reference value for determining whether the number of inputs for a specific search word is rapidly increasing. According to one embodiment of the present invention, the jump reference numerical information generating means 804 may generate the jump reference numerical information by adding the average value information multiplied by the standard deviation information by a predetermined number. For example, the predetermined number may be three. In this case, the sudden reference numerical information may be expressed by a formula such as "mean + (3 * standard deviation)".

The search word input means 805 receives a second search word from a user terminal, and the counter means 806 counts the number of inputs per time interval for each second search word to generate second input count information.

Comparing means 807 compares the second input count information and the spike reference numeric information, and the search word detecting means 808 is configured to be associated with third input count information that is equal to or higher than the spike reference numerical information among the second input count information. 3 Detect the search word.

As described above, the sudden rising search word detection system 800 compares the reference value based on the previous data of each search word with the number of inputs at the present time of each search word, thereby increasing the number of inputs at the present time. Search terms that record more than the reference value are detected as rising search terms. 4 is a diagram illustrating an example of a graph for detecting a rising search word through comparison with the rising reference numerical information as described above. 4 has already been described in the present specification, and thus a detailed description thereof will be omitted.

According to one embodiment of the present invention, there is provided a spike search word detection system capable of generating rank information of the detected spike search word, which will be described below.

The rapidly rising search word detection system 820 according to the present embodiment includes a second database 809, a second database management means 810, an input frequency increase information calculating means 811, an input ranking information retrieving means 812, and an input ranking. The rising information calculating means 813, the standard regular variable information generating means 814, the rising index information generating means 815, and the rising rank information generating means 816 may be further included.

The second database 809 includes search term input ranking information according to the first input count information. 5 is a diagram illustrating an example of the second database 809. Since FIG. 5 has already been described herein, a detailed description thereof will be omitted.

The second database managing means 810 generates the search word input ranking information at each time interval and records the search word input ranking information in the second database 809.

The input count increase information calculating means 811 refers to the database 801 to calculate the input count increase information associated with the third input count information.

The input ranking information retrieving means 812 retrieves the input ranking information corresponding to the third input count information by referring to the second database 809, and inputs the input rank increasing information calculating means 813 to the second database 809. The input rank increase information associated with the input rank information is calculated with reference to the figure.

The standard regular variable information generating unit 814 generates standard normal variable information about the third input frequency information by using the average value information and the standard deviation information.

The spike index information generating means 815 generates spike index information by using the number of times rising information, the input ranking information, the input ranking rising information, and the standard normal variable information.

According to an embodiment of the present invention, the spike index information generating means 815 generates a multiplication operation value of the input number of rising information, the reciprocal of the input ranking information, the input ranking rising information, and the standard normal variable information, and The surge index information may be generated using the multiplication operation value.

In addition, according to an embodiment of the present invention, the spike index information generating unit 815 may multiply the multiplication operation value of the input number of rising information, the reciprocal of the input ranking information, the input ranking rising information, and the standard normal variable information. In the generating, the method may generate order information associated with the standard normal variable information and generate the multiplication operation value using the order information. In this case, the order information may vary according to a predetermined criterion.

According to another embodiment of the present invention, the spike index information generating means 815 generates an addition operation value of the input number rising information, the input rank rising information, and the standard normal variable information, and adds the addition operation value and The multiplication operation value of the inverse of the input rank information may be generated, and the sudden increase index information may be generated using the multiplication operation value. Further, according to the present embodiment, the spike index information generating means 815 generates the first weight information associated with the input number rising information in generating the addition operation value, and the second index associated with the input rank rising information. Generate weight information, generate third weight information associated with the standard normal variable information, and generate the addition operation value using the first weight information, the second weight information, and the third weight information. Can be used. In this case, the first weight information, the second weight information, or the third weight information may vary according to a predetermined criterion.

The ranking information generating means 816 may generate the ranking information regarding the third search word based on the ranking index information.

The system operator may use the generated jump ranking information to provide a user with a more specific personalized service for a search word with a greater degree of spike and more important. In addition, the system operator can satisfy the user's intellectual desire to wonder what search terms are recently increasing in popularity by providing the generated rising ranking information to the user.

In the above, an embodiment of detecting a rising search word and ranking the detected rising search words has been described above. However, an embodiment of detecting the rising search word using the factors used to determine the ranking may exist. An example is demonstrated.

The rapidly rising search word detection system according to the present embodiment includes a database 801, a database management means 802, a second database 809, a second database management means 810, a search word input means 805, and a counter means 806. Input number rising information calculating means 811, input ranking information retrieving means 812, input ranking rising information calculating means 813, mean / standard deviation information generating means 803, standard normal variable information generating means 814 , Spike index information generation means 815, and search term detection means 808.

In the present embodiment, the search word detecting means 808 detects a rising search word based on the rising index information generated by the rising index information generating means 815.

According to the embodiment as described above, it is possible to detect the rising search word more accurately using a variety of factors. In addition, since the ranking information can also be generated by using the index information, there is an advantage that does not need to separately create the ranking information.

According to an embodiment of the present invention, 1) when the input rank increase information for the search word is a predetermined value or more, 2) when the search word newly enters within a predetermined rank according to the input rank for the search word, or 3) the search word. There is provided a system for calculating the input ranking rising information for a predetermined period and detecting the search word as a rising search word when the input ranking rising information is continuously a predetermined value or more during the period.

In the case of 1), the sudden rising search word detection system may include determining means (not shown) for determining whether the input rank rising information for the search word input from the user terminal is greater than or equal to a predetermined value. In addition, in the case of 2), the rising search word detection system includes determining means (not shown) for determining whether the search word is newly entered within a predetermined rank by using input ranking information of the search word input from the user terminal. can do.

On the other hand, in case of 3), the rising search word detection system calculates the input ranking rising information for calculating the input ranking rising information for the search word at each time interval in which the input count information is generated for a predetermined period with reference to the second database 809. Means 813 and determining means (not shown) for determining whether the input ranking increase information is continuously above a predetermined value during the period.

According to the embodiment 1) or 2), a search word whose input rank rises rapidly can be easily detected as a rising search word, thereby simplifying a process for detecting a rising search word, thereby reducing unnecessary memory requirements and The speed is improved. In addition, according to the embodiment of 3), a search word having a continuous ranking increase for a predetermined period of time can be detected as a rising search word, so that a search word that is continuously increasing in number instead of a search word that is temporarily increased in number of times is input. The effect can be obtained.

According to an embodiment of the present invention, a system capable of detecting a sudden rising search word by generating average value information and upper management limit information for each search word using an Exponentially Weighted Moving Average (EWMA) control chart, and determining whether the average value information is moved. Although provided, this embodiment will be described below. 9 is a block diagram illustrating a system for detecting a rising keyword according to the present embodiment.

The rapidly rising search word detection system 900 according to the present embodiment includes a database 901, a database management means 902, an average / upper limit information generation means 903, a search word input means 904, a counter means 905, and a determination. Means 906, and search term detection means 907.

The database 901 includes a search word identifier and first input count information for a first search word associated with the search word identifier, and the database managing means 902 generates the first input count information at predetermined time intervals. 901).

The average value / limit management upper limit information generating means 903 refers to the database 901 to obtain average value information and upper limit management information regarding the first input frequency information for each of the first search terms using an EWMA (Exponentially Weighted Moving Average) chart. Create

The search word input means 904 receives a second search word from a user terminal, and the counter means 905 generates second input count information by counting the number of inputs per time interval for each second search word.

The determining means 906 determines whether the average value information has moved based on the second input count information and the management upper limit information, and if it is determined that the average value information has moved, the search term detecting means 907 moves the movement. By detecting the third search word associated with the averaged value information, the rising search word may be detected.

Rising search word detection method and system according to the present invention can be used for various purposes useful.

For example, when a sudden rising search word is detected, the search service operator may designate it as a search word to be sold so that a search word advertisement sale may be performed through a predetermined keyword shop. In this case, the service operator may increase advertising revenue by widening the range of designating the rising search term as a search term and may maximize profit by adjusting a price of the search term if it is already being sold.

In addition, it is possible to combine the rising search word detection method and system according to the present invention to shopping. For example, the shopping mall operator may detect a rapidly rising product and post the introduction of the product on the front page of the shopping mall homepage, thereby maximizing the profits obtained by the shopping mall operator. In this case, the method of detecting a rapidly rising product may be applied to the method of detecting a rising keyword according to the present invention.

In addition, seasonal keyword can be extracted by analyzing the search word log data for several years using the rapidly rising search term detection method and system according to the present invention. For example, the search service operator extracts a search word that has a sharp increase in search requests in a certain period of time, and determines a popular period for each search term. Then, the search service operator adjusts the selling price of the search term immediately before the period, and also promotes the popular keyword. You can maximize your profits by strengthening.

In addition to the above, the method and system for detecting a rising search word according to the present invention may be utilized for useful purposes in various fields.

10 is an internal block diagram of a general purpose computer device that may be employed to perform the method for detecting spiked search words according to the present invention.

The computer device 1000 includes one or more processors 1010 connected to a main memory device including a random access memory (RAM) 1020 and a read only memory (ROM) 1030. The processor 1010 is also called a central processing unit (CPU). As is well known in the art, the ROM 1030 serves to transfer data and instructions to the CPU unidirectionally, and the RAM 1020 typically transmits data and instructions bidirectionally. Used to. RAM 1020 and ROM 1030 may include any suitable form of computer readable media. Mass storage 1040 is bi-directionally coupled to processor 1010 to provide additional data storage capability, and may be any of the computer readable recording media described above. The mass storage device 1040 is used to store programs, data, and the like, and is a secondary memory device such as a hard disk which is generally slower than the main memory device. Certain mass storage devices such as CD ROM 1060 may be used. The processor 1010 may include one or more input / output interfaces such as video monitors, trackballs, mice, keyboards, microphones, touchscreen displays, card readers, magnetic or paper tape readers, voice or handwriting readers, joysticks, or other known computer input / output devices. 1050 is connected. Finally, the processor 1010 may be connected to a wired or wireless communication network through the network interface 1070. Through this network connection, the procedure of the method described above can be performed. The apparatus and tools described above are well known to those skilled in the computer hardware and software arts.

The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention.

As described above, although the present invention has been described with reference to limited embodiments and drawings, the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined not only by the claims below, but also by the equivalents of the claims.

According to the method and system for detecting a rising search word according to the present invention, by generating a rising reference value by using the average value and standard deviation of each search word and comparing the rising reference value and the measured number of inputs to detect the rising search word In this case, an effect of providing a user with a customized service for the sudden search term may be obtained.

In addition, according to the method and system for detecting a rising search word according to the present invention, by quickly detecting a search word that the number of inputs from the user is rapidly provided to the user, a user who wants to quickly receive a search result using a rapidly rising search term Can achieve the effect of satisfying the needs of the.

In addition, according to the method and system for detecting the spiked search word according to the present invention, by quickly detecting a search word that is rapidly rising the number of inputs from the user, the Internet search service provider utilizes the detected spiked search term in various fields to obtain the revenue The effect of maximizing can be obtained.

In addition, according to the method and system for detecting the rising search word according to the present invention, by accumulating the data accumulated over all fields for a certain period of time to detect the rising search words, the user's desire to receive a more accurate and rising search terms in various fields The effect that can satisfy the can be obtained.

1 is a diagram illustrating a network connection of a fast rising keyword detection system according to an exemplary embodiment of the present invention.

2 is a flowchart illustrating a method of detecting a rising keyword according to an exemplary embodiment of the present invention.

3 is a view showing an example of a database in an embodiment of the present invention.

4 is a diagram illustrating an example of a graph for detecting a rising search word through comparison with a rising reference numerical information according to an embodiment of the present invention.

5 is a diagram for one example of a second database including search word input ranking information according to one embodiment of the present invention.

6 is a flowchart illustrating a method of detecting a rising search word according to another exemplary embodiment of the present invention.

7 is a flowchart illustrating a method of detecting a rising search word according to another exemplary embodiment of the present invention.

8 is a block diagram illustrating a system for detecting a rising keyword according to another embodiment of the present invention.

9 is a block diagram illustrating a system for detecting a rising keyword according to another exemplary embodiment of the present invention.

10 is an internal block diagram of a general purpose computer device that may be employed to perform the method for detecting spiked search words according to the present invention.

<Explanation of symbols for main parts of the drawings>

800: rising search term detection system

801: Database 802: database management means

803: mean / standard deviation information generating means

804: Sudden increase numerical information generating means

805: search term input means 806: counter means

807: comparison means 808: search term detection means

Claims (1)

In the method for detecting a search word that rises rapidly the number of inputs from the user terminal, Maintaining a first database comprising a search term identifier and first input count information for a first search term associated with the search term identifier, wherein the first input count information is generated at predetermined time intervals and recorded in the first database; ; Maintaining a second database including search term input ranking information according to the first input count information, wherein the search term input ranking information is generated at each time interval and recorded in the second database; Receiving a second search word from a user terminal; Generating second input number information by counting the number of times of input per time interval for each second search word; Searching input ranking information corresponding to the second input count information by referring to the second database; Calculating first input rank increase information associated with the input rank information by referring to the second database; Determining whether the first input rank increase information is equal to or greater than a predetermined value; And Detecting a third search word associated with second input rank rising information that is equal to or greater than the numerical value among the first input rank rising information based on the determination result; Including, Rapid search word detection method characterized in that for changing the numerical value according to the input ranking information.