WO2013091302A1

WO2013091302A1 - Intelligent search method and device

Info

Publication number: WO2013091302A1
Application number: PCT/CN2012/071663
Authority: WO
Inventors: 陈伟
Original assignee: Chen Wei
Priority date: 2011-12-22
Filing date: 2012-02-27
Publication date: 2013-06-27
Also published as: CN102385636A

Abstract

Disclosed are an intelligent search method and device. The method includes: step 1, acquiring current location information about a user; step 2, searching with a customized search model corresponding to the user to obtain point of interest information and/or event information about the user according to the current location information and the current time, wherein the customized search model is obtained by performing demand analysis on the basic population to which the user belongs in each time period; and step 3, pushing the point of interest information and/or event information to the user. The present invention can obtain point of interest information and event information based on user features without user input operations and can conveniently and accurately meet the daily-life information demand of the user.

Description

Intelligent search method and device

The present invention relates to intelligent search technology, and in particular to an intelligent search method and apparatus. Background technique

In today's society, the Internet has become popular, which brings a huge amount of information. In the vast ocean of information, people can only find the information they need by relying on the search engine (s earch eng ine). The so-called search engine refers to a system that collects information from the Internet according to a certain strategy, uses a specific computer program, and provides a search service for the user after organizing and processing the information, and presents the user with relevant information to the user.

The intelligent search engine is a new generation of search engines that combine artificial intelligence technology. In addition to providing traditional fast retrieval and relevance sorting functions, he also provides functions such as user role registration, automatic identification of user interests, semantic understanding of content, intelligent information filtering and push.

The existing intelligent search engine technologies mainly include:

(1) Keyword search and its derived semantic search are the technical means for existing users to obtain information. This technique relies on the user's input and then performs a relevant search in the system database and then provides the results to the user.

The main disadvantages of keyword search and its derived semantic search are:

1. This technology requires the user to input keywords or words manually or by voice, with slow input on the mobile device, prone to errors, and so on.

2. This technology requires keywords and words, but some user requirements cannot be expressed in a limited key.

(2) Data acquisition technology based on the geographic location of the user. This technique, by taking the user's geographic location information, then retrieving it in the database, then presents the results to the user in terms of consumption categories or distances.

The disadvantages of data acquisition techniques based on user geography are:

1, just a simple list of search results, can not be based on time, events, user characteristics and other factors to determine the user's true intention to display.

2. Users need to spend time browsing the information they get and filter by themselves, which is not efficient. Summary of the invention

An object of the present invention is to provide an intelligent search method and device, which can obtain interest point information and event information based on the user's own characteristics without user input operation, and can conveniently and accurately meet the user's life information requirements.

In order to achieve the above object, the present invention provides an intelligent search method, including: Step 1: acquiring current location information of a user;

Step 2: According to the current location information and the current time, perform a search by using a customized search model corresponding to the user, and obtain interest point information and/or event information for the user; where the customized search model is The basic group to which the user belongs is obtained by performing demand analysis for each time period;

Step 3: Push the interest point information and/or event information to the user.

Preferably, in the above method, the customized search model includes:

Calculating the attention degree of the demand according to the basic population to which the user belongs, the time of the establishment is the horizontal axis, the attention degree is the demand concern curve of the vertical axis, and the demand attention curve according to different requirements obtains the current time point according to the level of the attention degree. Order of demand;

Various requirements are classified and defined, and corresponding to the points of interest.

Preferably, in the above method,

Determining a basic population to which the user belongs according to a user characteristic of the user;

The user features include: age parameters, gender parameters, occupation parameters, and/or preference parameters obtained by means of user registration, usage analysis, question and answer interaction, active addition, and the like.

Preferably, the method further includes: modifying, by the interest point region influence model, the customized search model, wherein the interest point region influence force model comprises: determining an attenuation radius of the influence according to the level and type of the interest point And the attenuation curve, the degree of attention of the demand and the interest point is corrected according to the distance between the point of interest and the current position of the user.

Preferably, in the above method, the method further includes: modifying, by the disturbance factor model, the customized search model, the disturbance factor model comprising: correcting the degree of attention of the demand and the interest point according to the objective event.

Preferably, in the foregoing method, the objective event includes a preset known event and an acquired emergency event;

The preset known events are: season, seasonal, solar, festival, and/or local customs; the acquired emergencies are: weather, traffic conditions, social news events, and/or tides Stream hotspots.

Preferably, in the above method, the method further includes: modifying, by the daily accumulation, the frequency and appearance time period of the user at a specific location or a point of interest location, and modifying the customized search model.

Preferably, in the foregoing method, the step 3 further includes: determining, according to user characteristics of the user, a display type, a display quantity, and/or an arrangement order of the interest point information and/or event information;

The point of interest information includes: a name, a category, a longitude, a latitude, and a phone call; the point of interest information further includes: an experience comment, a dynamic offer, an encyclopedia knowledge, and/or a topic;

The event information is: traffic conditions, social news events, and/or hotspots.

The invention also provides an intelligent search device, comprising:

An information obtaining module, configured to: obtain current location information of the user;

a search module, configured to: obtain, according to the current location information and the current time, a point of interest and/or event information for the user by performing a search by using a customized search model of the user; wherein the customized search model It is obtained by performing a demand analysis on each time period of the basic population to which the user belongs;

The pushing module is configured to: push the point of interest information and/or event information to the user.

Preferably, in the foregoing apparatus, the information acquiring module is further configured to: obtain, according to the user's knowledge, the user's age parameter, gender parameter, occupation parameter, and the like by user registration, use analysis, question and answer interaction, active addition, and the like. / or preference parameters; on the premise of the user's knowledge, the user's time and user identification number are obtained when the user logs in or refreshes the software;

The search module includes: a modeling module, configured to: calculate a degree of attention of a demand according to a basic population to which the user belongs, establish a time-sharing curve with a horizontal axis, and a focus on a vertical axis, and pay attention to the demand according to different needs The curve obtains the order of demand according to the level of attention value at the current time point; the requirement code hooking module is used to: classify various requirements and define corresponding relationships with the points of interest;

The modeling module is further configured to: modify the customized search model by using a point of interest regional influence model, and modify the customized search model by using a disturbance factor model, by using the user to be in a specific location and interest The frequency and appearance of the point location a time period, modifying the customized search model;

The push module is further configured to: determine, according to a user feature of the user, a display category, a display number, and/or an arrangement order of the interest point information and/or event information; the interest point information includes: a name, a category , longitude, latitude, and telephone; the point of interest information also includes: experience reviews, dynamic offers, encyclopedic knowledge, and/or topics.

At least the following technical effects exist in the embodiments of the present invention:

1) In the embodiment of the present invention, the user can obtain the interest point information and the event information based on the location, the time, and the self-characteristics without inputting the keyword, thereby solving the prior art input trouble, the keyword inaccuracy, and the user information acquisition being inaccurate. Such defects can help users to obtain life information and services conveniently and efficiently.

2) The embodiment of the present invention is a technology for continuously accumulating behavior characteristics of each user, establishing a personalized demand model for the user, and providing life service information according to the demand, and is not passively waiting for the user to make a keyword search request. Instead, it conveniently pushes the most relevant life service information to the user based on location, time, character model and characteristics.

In summary, the embodiment of the present invention can maximize the real needs of the user through user modeling and requirement analysis, and provide the user with information of interest to the user, which greatly improves the efficiency of user information acquisition. DRAWINGS

1 is a flow chart of steps of an embodiment of a method according to the present invention;

2 is a structural diagram of a smart search device provided by the present invention;

Figure 3 is a demand attention curve provided by the present invention;

FIG. 4 is a schematic diagram of a demand sequence arranged according to the degree of attention provided by the present invention. detailed description

The specific embodiments are described in detail below with reference to the accompanying drawings.

1 is a flow chart of the steps of the method embodiment of the present invention. As shown in FIG. 1, the embodiment of the present invention provides an intelligent search method, which includes:

Step 101: Obtain current location information of the user.

Step 102: Perform a search by using a customized search model corresponding to the user according to the current location information and the current time, and obtain information about interest points and/or things for the user. And the customized search model is obtained by performing a demand analysis on each time period of the basic population to which the user belongs;

Step 103: Push the interest point information and/or event information to the user. It can be seen that the embodiment of the present invention performs the search by using the current location and the customized search model, and the user can obtain the interest point information or the event information based on the self-characteristics without inputting the keyword, so that the input trouble of the prior art and the keyword inaccuracy can be solved. Defects such as inaccurate user information acquisition, which can help users to obtain life information and services conveniently and efficiently.

The customized search model includes: calculating a demand attention degree according to the basic population to which the user belongs, establishing a demand attention curve with a horizontal axis and a vertical axis of interest, and forming a concern for various needs based on the basis The order of the degree values; the various requirements are classified and defined, and the points of interest are associated with each other.

FIG. 3 is a demand attention curve provided by the present invention, as shown in FIG. 3. It shows the change in movie demand for a certain user (for example, a freelance woman) on a certain weekend (for example, December 16, 2011). Among them, the time is the horizontal axis, and the attention of the movie demand is the vertical axis. It can be seen that the attention of the movie from 16 to 18 is relatively large, and this time is most concerned about the video and is easier to make a decision to go to the movie. The attention paid to watching movies in the early hours of the morning is very small, because people are sleeping during this time.

FIG. 4 is a schematic diagram of a demand sequence arranged according to the degree of attention provided by the present invention. As shown in the figure, the demand curve of various users (for example, freelance women) is sliced, and the attention of various needs at a certain point in time can be obtained, and after being ranked according to the degree of attention, It is known that at this point in time, the user's greatest demand is those. For example, in Figure 4 is the order of demand at 6 o'clock in the afternoon. At this point in time, the user's greatest demand is eating, watching movies and shopping.

The basic population to which the user belongs is determined according to the user characteristics of the user; the user features include: age parameters, gender parameters, occupation parameters, and/or obtained through user registration, usage analysis, question and answer interaction, active addition, and the like. Or hobby parameters.

The method further includes correcting the customized search model by daily occurrence of the frequency and appearance time period of the user at a specific location or point of interest location.

Therefore, the embodiment of the present invention is a technology for continuously accumulating behavior characteristics of each user, establishing a personalized demand model for the user, and providing life service information according to the demand, and is not passively waiting for the user to make a keyword search request. Instead, it conveniently pushes the most relevant life service information to the user based on location, time, character model and characteristics. In the present invention, not only the customized search model for the user is obtained by analyzing and modeling the user's needs, but also the interest point itself and the influence of the external conditions on the demand are modeled, and the influence model and the disturbance factor model of the interest point are obtained. Therefore, the user's needs can be grasped more accurately.

Therefore, in the embodiment of the present invention, the method further includes: modifying, by the interest point region influence model, the custom search model, wherein: the interest point region influence force model comprises: determining an attenuation radius of the influence according to the level and type of the interest point And the attenuation curve, the degree of interest of the point of interest is corrected according to the distance between the point of interest and the current position of the user.

For example, the user generally refuels nearby, and the distance is far greater than the user's influence. Therefore, the demand level of the gas station will decay rapidly with the extension of the distance, and the attenuation curve is steep and the attenuation radius is small.

For example, for a visitor to Beijing for the first time, although it is far away from some attractions, the famous attractions such as Tiananmen Square, Summer Palace and Great Wall still have great attention even if they are not nearby. .

In the embodiment of the present invention, the method further includes: modifying the customized search model by using a disturbance factor model, where the disturbance factor model comprises: correcting the degree of interest of the interest point according to the objective event.

The objective event includes a preset known event and an acquired emergency event; the preset known events are: season, seasonal, solar, festival, and/or local customs; the acquired emergency For: weather, traffic conditions, social news events and/or hotspots.

For example, in the Dragon Boat Festival, the attention of restaurants that can eat dumplings will increase, and in winter, the attention of cold drink shops will decline. During the SARS epidemic, the attention of pharmacies and hospitals will increase.

In addition, the present invention does not simply push the blunt point of interest information to the user. The step 103 further includes: determining, according to the user feature of the user, the display type, the number of displays, and/or the sorting order of the point of interest information. The interest point information includes: name, category, longitude and latitude, telephone, and may additionally include: experience reviews, dynamic offers, encyclopedia knowledge and/or topics. The event information is: traffic conditions, social news events, and/or trend hotspots

2 is a structural diagram of a smart search device provided by the present invention. As shown in FIG. 2, the smart search device includes: The information obtaining module 201 is configured to: obtain current location information of the user; the searching module 202 is configured to: perform, according to the current location information and the current time, search by using a customized search model corresponding to the user, to obtain a target for the user a point of interest information and/or event information; wherein the customized search model is obtained by performing a demand analysis on each time period of the basic population to which the user belongs;

The pushing module 203 is configured to: push the point of interest information and/or event information to the user.

The search module is further configured to: modify the customized search model by using a point of interest regional influence model, and modify the customized search model by using a disturbance factor model, by using the daily accumulated user at a specific location, The custom search model is modified by the frequency of occurrence and the time period of occurrence of the point of interest location.

It can be seen that the modeling models used in the intelligent search device mainly include the following:

1) Corresponding user's customized search model, including character feature modeling and character basic requirement modeling.

Character Feature Modeling: It gradually accumulates user characteristics based on user interaction, usage preferences and data analysis, and keyword maintenance, and determines the impact value for various demand concerns.

Character basic needs modeling: classify and encode human and consumer-related basic needs, forming dozens, hundreds or thousands of demand code tables. According to the gender, place of residence, time and freedom, etc., the time period of demand of different people is compiled, and the demand concern curve with time as the horizontal axis and attention as the vertical axis is established.

2) The interest point regional influence model, that is, POI (Point of Interes t interest point) and regional influence modeling

According to the requirements coding and POI characteristics, it implements the hanging code operation on the specific POI and the area, and determines its impact value on the demand; determines the influence attenuation radius according to the POI level; and calculates the influence field of the POI in each area of the city. .

3) Disturbance factor modeling

Set the impact on various demand focus curves based on predictable objective facts (such as seasons, seasonality, solar terms, festivals, local customs, etc.) and unpredictable objective facts (such as weather, traffic conditions, social events, hotspots, etc.) value.

The above model constitutes the core engine of intelligent search, and superimposes the above demand attention value and its influencing factors to obtain the demand attention sequence of the user at a specific time in a specific time period. In addition, the push module 203 has the following information display rules: the user sends geographic location information, time point information, and automatically pushes to the server (ie, the smart search device of the present invention) for processing; the server generates a concern according to the core engine to generate a specific time-specific location. Sequence; then call out specific information from the content database based on the information display rules. The information display rules vary greatly depending on the type of person and whether they are resident. The main requirements are the type of display, the number of displays, and the order of arrangement. The content is mainly presented in the form of experience reviews, dynamic offers, and thematic encyclopedias, not a blunt POI profile.

As can be seen from the above, the embodiments of the present invention have the following advantages:

1) In the embodiment of the present invention, the user can obtain the interest point information and the event information based on the self-characteristics without inputting the keyword, thereby solving the defects of the prior art, such as input trouble, inaccurate keyword, inaccurate user information acquisition, and the like. It can help users get life information and services conveniently and efficiently.

In summary, the embodiment of the present invention can maximize the real needs of the user through user modeling and requirement analysis, and provide the user with information of interest to the user, which greatly improves the efficiency of user information acquisition.

The above description is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. It should be considered as the scope of protection of the present invention.

Claims

Rights request

1. An intelligent search method, comprising:

Step one: obtaining current location information of the user;

2. The intelligent search method according to claim 1, wherein the customized search model comprises:

The intelligent search method according to claim 2, wherein the basic population to which the user belongs is determined according to the user characteristics of the user; the user features include: through user registration, usage analysis, question and answer interaction, and initiative Add the age parameter, gender parameter, occupation parameter and/or preference parameter obtained by the method.

The smart search method according to claim 2, further comprising: modifying the customized search model by a point of interest regional influence model, wherein the interest point region influence model comprises: according to the point of interest The level and type determine the attenuation radius and attenuation curve of the influence, and correct the attention of the demand and interest points according to the distance between the point of interest and the current position of the user.

5. The intelligent search method according to claim 2, further comprising: modifying the customized search model by a disturbance factor model, the disturbance factor model comprising: correcting the attention of the demand and the interest point according to the objective event degree.

6. The intelligent search method according to claim 5, wherein The objective events include preset known events and acquired emergencies;

The preset known events are: season, seasonal, solar, festival and/or local customs;

The acquired emergencies are: weather, traffic conditions, social news events, and/or hotspots.

7. The intelligent search method according to claim 2, further comprising: correcting the customized search model by daily occurrence frequency and occurrence time period of the user at a specific location or a point of interest location. .

The smart search method according to claim 3, wherein the step 3 further comprises: determining, according to the user characteristics of the user, a display type and a display bar of the interest point information and/or event information. Number and/or order;

The point of interest information includes: name, category, longitude, latitude, and telephone;

The point of interest information also includes: experience reviews, dynamic offers, encyclopedia knowledge and/or topics;

The event information is: traffic conditions, social news events related to consumption, and/or hotspots.

9. An intelligent search device, comprising:

An information obtaining module, configured to: obtain a current location information of the user, and a search module, configured to: search, according to the current location information and the current time, by using a customized search model corresponding to the user, to obtain a point of interest for the user Information and/or event information; wherein, the customized search model is obtained by performing demand analysis on each time period of the basic population to which the user belongs;

a pushing module, configured to: push the point of interest information and/or event information to the user.

The intelligent search device according to claim 9, wherein the information obtaining module is further configured to: obtain the user by means of user registration, use analysis, question and answer interaction, active addition, etc., on the premise of the user's knowledge; Age parameter, gender parameter, occupational parameter and/or preference parameter; before the user knows The user's time and user identification number are obtained when the user logs in or refreshes the software;

The search module includes: a modeling module, configured to: calculate a degree of attention of a demand according to a basic population to which the user belongs, establish a time-sharing curve with a horizontal axis, and a focus on a vertical axis, and pay attention to the demand according to different needs The curve obtains the order of demand according to the level of attention value at the current time point; the requirement code hooking module is used to: classify various requirements and define a correspondence relationship with the points of interest;

The modeling module is further configured to: modify the customized search model by using a point of interest regional influence model, and modify the customized search model by using a disturbance factor model, by using the user to be in a specific location and interest Modifying the customized search model by the appearance frequency and the appearance time period of the point location; the pushing module is further configured to: determine, according to the user characteristics of the user, a display type of the interest point information and/or event information, Displaying the number of bars and/or sorting order; the points of interest information includes: a name, a category, a longitude, a latitude, and a phone call; the point of interest information further includes: an experience comment, a dynamic offer, an encyclopedic knowledge, and/or a topic.