WO2010111895A1 - System and method for remote feature consulting - Google Patents

System and method for remote feature consulting Download PDF

Info

Publication number
WO2010111895A1
WO2010111895A1 PCT/CN2010/000431 CN2010000431W WO2010111895A1 WO 2010111895 A1 WO2010111895 A1 WO 2010111895A1 CN 2010000431 W CN2010000431 W CN 2010000431W WO 2010111895 A1 WO2010111895 A1 WO 2010111895A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
group
plurality
individual
feature
Prior art date
Application number
PCT/CN2010/000431
Other languages
French (fr)
Chinese (zh)
Inventor
范晓
Original Assignee
Fan Xiao
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to CN 200910048809 priority Critical patent/CN101853428A/en
Priority to CN200910048809.0 priority
Application filed by Fan Xiao filed Critical Fan Xiao
Publication of WO2010111895A1 publication Critical patent/WO2010111895A1/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services

Abstract

A remote feature consulting system for consulting status of feature is provided, which includes an input device for collecting input feature data which includes an individual identifier item and individual identifier item data, multiple feature items and feature item data corresponding to each feature item; and a remote server which includes a storage device, an individual feature analysis device, a group feature analysis device and a comprehensive feature analysis device, therein the storage device stores the feature data according to the individual identifier item data in the feature data. A corresponding remote feature consulting method for consulting status of feature is also provided.

Description

Features and remote advisory system BACKGROUND

The present invention relates to a remote system and remote consulting advice features methods. Background technique

Currently, dedicated to fast, accurate and offers a variety of personal information and corporate activities rapid rise analysis guidance counseling services. Furthermore, with the rapid development of information processing equipment such as a computer's performance and remote computer systems, capable of large-scale data processing of remote consultation system based on a remote computer system is increasingly becoming an important means of providing information consulting services consulting service industry.

One application of remote consultation system is to collect information on the various features of an individual or population, such as health information, shopping habits and other information, and feature remote Consultation corresponding analysis, such as remote health advisory system, remote shopping habits of consultation system. These advisory systems feature the results of analysis of the individual's life or business activities is very meaningful. Therefore, people have been working with concurrency features such as remote consulting system.

For used in personal / population health in the field of remote health advisory system, since the 1980s there have been a variety of remote health advisory system gradually come out and put into practical use, more and more users willing to use such a remote health advisory system access to a variety of health information, while eliminating the need for time in the hospital often suffer from waste, and other costly problems. For the analysis of individual / population shopping habits remote shopping habits advisory system, analyze the information these systems can provide for a large supermarket or store operations is very significant. However, the prior art remote health advisory system, remote shopping habits of consultation system still has many deficiencies. From a technical point of view, the conventional remote feature, such as health ^ inquiry system has many deficiencies in terms of a system ¾ meter, the data, e.g., lack of high-level analysis process functional group analysis and personal data such as a combination of analysis, situation and the lack of features, functions, and so health conditions such as predictive analysis process. , Summary

View of the above deficiencies of the prior art made in the present invention. Object of the present invention is to provide a remote consultation system characteristics, characterized in that the remote system can consult a comprehensive analysis process including individual analysis processing, analysis processing and group of individuals and groups for combining the personal characteristic data collected, and can be corresponding predictive analysis process, the analysis result can be provided with higher accuracy.

According to one aspect of the present invention, there is provided a system for remote consultation characteristic feature status for consulting, characterized by comprising: feature data input means, collecting input data comprises an individual identifier and individual identifier item items data, wherein a number of data items and the characteristic feature item corresponding to the item; and a remote server comprising storage means according to the individual identifier of the project characteristic data of data stored feature data; wherein the individual analysis apparatus, according to the storage means storing specific individual ^^ views and features of currently input data to generate the individual results of statistical analysis and the individual correlation results and, based on the particular individual previous and the current input feature data, the individual results of statistical analysis and the individual correlation results generates individual characteristic analysis jobs taken report, the group group wherein analyzing means generating a first plurality of groups based on all the previous individual storage device and the feature data stored in the input current, a first plurality of groups corresponding to the statistical analysis group and the groups associated with the results, and based on a first plurality of groups, group Statistical analysis of the group and group association result generating analysis feature, feature analysis and synthesis means, a first means for generating a plurality of groups according to a specific analysis of individual storage means stores the previous and current feature groups and feature data input comprehensive analysis of the previous input feature generation particular individual characteristic data corresponding to a second plurality of results of predictive analysis and to a second group of a plurality of groups of optimization, and generates a plurality of a second analysis result based on the prediction and optimization of the group report.

† According to another aspect of the present invention, there is provided a method for remote consultation characteristic feature status for advice, characterized by comprising: an input process, input feature data collection, data including a feature identifier for the item and the individual identification sister character data item, and the number of feature item corresponding to the item of the characteristics of the feature data items; storage process, the identifier depending on the individual items stored in the feature data in the feature data transactions; individual characteristics analysis process, according to the particular individual input current and previous Tau feature data generation and statistical analysis of a subject associated with the node table and the data based on a particular subject and the current input the previous features, statistical analysis and individual subject ^ off individual characteristics result generating analysis, group analysis processing wherein, based on all the individual wherein the current input data and previous generates a first plurality of groups, the first group of the plurality of groups and the statistical results Fang correlation result corresponding to the group, based on a first plurality of groups, group statistical analysis and wherein groups associated group result generating analysis reports, and synthesis Zheng analysis process, the analysis processing of generating the feature data to generate a first plurality of groups corresponding to a particular individual previous input group and to a second plurality of second multi According to a particular individual and previous input feature data and the current group wherein optimization of results of predictive analysis of groups, and generating a comprehensive analysis wherein a second analysis result based on the prediction and optimization of the plurality of groups. BRIEF DESCRIPTION

. Incorporated in and constitute a part of the specification with drawings for explaining the present invention. According to the embodiment, and the detailed description of the foregoing general description and the following embodiments with ¾ of the principles of the invention described. .

Figure 1 is a block diagram showing an exemplary overall configuration of a remote health advisory system of the first embodiment according to the invention ^; Figures 2A and 2B is based on health data processing remote health advisory system embodiment of the present invention. a schematic view of an example;

FIG 3 is a block diagram illustrating an exemplary Ju made health analysis apparatus according to the individual remote health car advice system of the first embodiment of the present invention;

FIG 4 is a block diagram showing an exemplary configuration of the apparatus according to the group of the Healthy remote health advice system of the first embodiment of the present invention;

FIG 5 is a block diagram illustrating an exemplary configuration of a remote health according to general health advice system of the first embodiment of the present invention, the analysis apparatus;

6A is a schematic of an exemplary generation process sacrifice predictor of risk for hypertension consultants analyze an example display;

6B is an exemplary predictor for the risk of hypertension analysis consultant example; FIG. 6C is a prediction for a plurality of different groups of risk analysis of hypertension Illustrative examples; FIG. 7 is a block diagram illustrating an exemplary overall configuration of a remote health advisory system of the first modification of the first embodiment of the present invention according to the display;

Is a detailed system design according to a first remote health advisory system of the second embodiment variant of the embodiment of the present invention, lb;

FIG 9 is a diagram illustrating connection provider remote health advisory system according to a second modification of the first embodiment of the present invention with an individual lb health advisory service usage;

FIG 10 is a block diagram showing an overall configuration of an exemplary process ¾ shopping habits consultation system to a second embodiment of the present invention; to ¾

FIG 11 is a shopping cart of the processing in accordance with customary remote consultation system of the second embodiment of the present invention: a schematic view of an example of data used to ¾. . detailed description

It will now be described with reference to the accompanying drawings of various embodiments of the present invention.

(First Embodiment)

1 will now be specifically described with reference to FIG remote health advisory system as a first embodiment of the process wherein δ advisory system of the present invention 1, wherein the object feature refers to certain characteristics or events. "Health" is a feature of with: ίΦ statements. FIG. 1 is a block diagram showing an exemplary overall configuration of a remote health advisory system according to the first embodiment of the present invention. Remote health advisory system is a system of status as a characteristic example of the state of health counseling. Characteristics situation in which specific performance characteristics as an individual. As a characteristic example of the state of health is the health of specific performance, including, ^ disease, such as hypertension, heart disease, headaches, stomach aches, as well as sub-health physical condition, normal physical condition, physical condition and other athletes or health status levels Wait.

In the remote health advisory system as a first embodiment of the remote consultation system characteristic of the present invention, the object of processing is an individual (i.e., user) health data, health data is an example of the characteristic data of the present invention. Individual article can be individuals, it can also be a family or units. The following will first be described in detail as an example of the health data feature data.: -

'Personal health consult remote health advisory system according to a first embodiment of the present invention may be applied, it may be applied by a number of the health of the population of individuals for consultation. 1 of the following situations apply to consulting in remote personal health counseling system health (that is, the individual is personal) health data are processed first described. 2A and FIG. 2B shows an example of a remote health advisory system of the first embodiment of the present invention a process of health data. Health data and project data including the items corresponding to each item. Wherein the items refer to items healthy individuals as an identifier for the item and a plurality of feature item instance. Examples of health data items as feature data items corresponding to the specific data refers to health programs, which represents a numerical value or a code. Health data may further include a data input date (health data input time), the time input of health data is an example of feature data input time, represents the input time for each of the health data. 2A and 2B are respectively input time instance 2009-12-1 and 2009-12-3 two different health data. Although not shown, if the process to be considered in the case of aging, the health data may include time data generated by various health programs, i.e. each data item health adding time which may be the health data input time, for example, "today drinking habits: Y <2009-12-1>". It should be noted that, where the health program "Drinking habits today" is just an example as to "Day" is entered in a health data shows that when health data on a "week," as a unit, or a "specific time" for the the unit, set up health programs to do the appropriate changes, such as changes to "this week whether drinking" ', "this time whether alcohol consumption."

In health data, the individual project identifier is used to distinguish individual projects. As shown in FIG 2Α and 2Β an example, the individual identifier of each data item corresponding to the item identifier includes individual name, ID number, mobile telephone number.

In the health data, wherein a portion of the health data items are able to characterize the direct health data, such as height and weight ratio, blood pressure, blood oxygen content, etc., as well as health data item is the data portion may have an impact on the health status, including age, living conditions (such as occupation, late at night, exercise, alcohol consumption, etc.), family status (marital status, number of children) and so on.

In the health data, health programs present a variety of modes. Health respective items may exist either mode to the standard mode of questionnaire items instrument measurement items, present in the health model instrument measurements which projects through a dedicated data anthropometric instruments from the project, there will be the following model instrument measurements of health programs referred to as instruments to measure health programs, heartbeat, blood pressure and other projects such as. Health and measuring instrument corresponding to the item instrument for measuring health data items, such as blood pressure, blood oxygen content. Present in the standard mode health questionnaire items is a standard item questionnaire remote health advisory system 12a of the input apparatus 1 provided to an individual item, the following modes will be present in a standard questionnaire item referred to as a standard item health questionnaire health programs, such as the user's age, living conditions (such as late at night, exercise, alcohol consumption, occupation, etc.), family status (number of children, marital status) and so on. Standard health questionnaire and standard questionnaire health project data corresponding to the item, such as age, occupation code, etc.

It is noted that the process can be carried out in accordance with the embodiment of the present invention requires the presence of the respective mode flexible health programs provided, for example, project from a healthy human body measurements, height, weight, etc., may be provided as a standard item health questionnaire. The type and number of healthy items may vary depending on usage, these are apparent to those skilled in the art.

Further, in the embodiment of the present invention, to make the description simpler, assume health program the instrument in the measurement data of health project data directly characterizing health data standard questionnaire Health Initiative data may have an impact on the health status, 2A and 2B. And health programs do not have to be in practical use is assumed in accordance with this setting, such as a standard item questionnaire of health data may also include data characterizing health directly, such as height, weight, etc., it will be apparent to those skilled in the art.

Described above is the case of the introduction of personal health counseling 1 used in remote health systems in the process of health data. When you talk to a remote health advisory system 1 is applied to the health of the population, in a remote health advisory system a health data processing and case consultation for personal health slightly different. Health data remote health advisory system one dealing with a crowd composed of more than similar situation, such as a home or a unit (ie, individual) health-related data. Here, "similar situation" means that they may belong to the same profession, or a member of lifestyle similar to the same family, and so on. In the case of remote health consultation Health consultation system 1 is applied to the crowd, that the individual (ie a crowd) health data is actually integrated a number of personal health data. In such health data, the same individual comprising an individual identifier to distinguish between different items, such as account number, phone number, etc., further comprising a health programs, including a direct health characterize individual items (i.e., populations) of overall health, and possibly have an impact on the health of health programs, namely health projects related to individuals (ie people) living conditions, which should be noted that some health project data directly characterize the state of health should be through directly reflect the appropriate post processing the individual (ie population) data overall health of more than one person.

Next, referring back to FIG. 1 illustrates the various components of a remote health advisory system. Remote health advisory system 1 includes a remote server 10, an input device and an output device 12a 12b, further comprising a storage means 140, 110 healthy individuals, the group health analysis means 120 and analyzing means 130 analyzes the integrated device 10 in the remote server.

Role of the input device 12a is input to collect health data. The input device transmits the individual health data to a remote server 10 for processing. When the remote server 10 completes the processing for health data and generate reports health analysis, and an output device 12b receives a predetermined format to output health analysis report to the user. Wherein the analysis includes an individual healthy healthy individual characteristics analysis as reported in example analysis, general health analysis Group health analysis report as an example of a group of feature analysis and an example of an integrated feature of the analysis, these reports the contents include the diagnosis, the health of predictive analysis, evaluation or grading health recommendations for improvement of the health status and so on.

Any hardware capable of exchanging data with the remote server 10 based on the input devices 12a and 12b in the output device 1 remote health advice system of the first embodiment of the present invention, software, or a combination of hardware and software. For example, input devices 12a and 12b may be an output device include a portable / desktop computers, mobile phone, PDA, or the like specified device consulting service provided in the. It should also be appreciated that the input means and output means 12a 12b may be present in a common apparatus, for example, by the same mobile phone, PDA, portable computer or the like, to collect health data to complete both health and complete analysis of the output, or test, can also be present separately in different apparatus, for example Mu collect health data through the portable computer, mobile phone and rely output health analysis.

In the process embodiment of the present invention, the input means 12a and output ^ counter 12b is set to automatically (according to the prevailing communication connection environment) or manually (user preferences) to select different transmission schemes, such as those based FTP, HTTP protocol documents transmission or based on GSM, GPRS message transmission protocol and the like. Means 12a and 12b for data transmission between the remote server 10 may need to be taken in accordance with various multiple transmission rules, such as the timing / immediate transmission rules, the encryption / cleartext rules.

Each remote server apparatus 10 is included in the remote health advice system 1 core member for processing health data. Next, they were described in detail.

1. The memory device 140

140 from the input device 10 stores the remote server apparatus 12a receives the health data stored for other devices that the remote server 10. When storing the health data, the storage device 140 may store the health data based on the individual health data item identifier data, to the health data belong to different hand differentiate individuals. For example, the storage device 140 can rely on an individual item identification data of the health data, a name, ID number, mobile phone number or the like in any combination thereof, such as to generate a unique identification to distinguish the individual subject ή substituting Shima, and divided in accordance with the individual identification code storage area in the storage device 140. Health of an individual belonging to the same data in the same storage area 140, it has been to achieve healthy individuals to distinguish between different data stored in the memory. Or a different individual of the health data are mixedly stored in the storage area, each having a unique health index data, and then to the individual identification code for each individual to establish a dedicated index table to point to all of the health data belonging to the individual. It may also be stored in the storage device 14 to the health data belonging to the same individual health data together in any suitable manner familiar to the person skilled in the art. The storage area may be a different storage device (such as an HDD), it may be the same storage device (HDD) in different partitions may be different databases.

In the case where the user has repeatedly used the ¾ Cheng Jiankang consultation system 1, the storage device 140 based on the identification code is stored with a previous use of the individual remote health advisory system health data input 1. When the individual, i.e., a new user input current health data, the storage device 140 according to the individual identification code of the determined position of the current input data and the database of the health of the individual any original previous input health data together. When healthy individuals without current input data, the current input of health data processing as a zero value.

When the health data stored in the storage means 140 in the remote server 10, the individual remote health analysis server device 10110, the group health and general health analysis apparatus 120 will be called the analyzer 130 from the storage device 140 as needed need of health data processing.

2. Analysis of individual health apparatus 110 generates individual subject and statistical analysis according to a particular individual associated with previous results stored in the storage means 140 and the current health data, and based on a particular calendar 110 individuals' health of an individual analysis apparatus ^: Current and input health data, the statistical analysis of individual and individual healthy individual correlation analysis result generating analysis reports as an example of individual characteristics. Healthy individual as the individual analysis apparatus Laid ¾Ε example analysis apparatus in ίθ remote server 110 described in detail with reference to the following FIG. FIG 3 is a block diagram illustrating an exemplary configuration of the health of an individual analysis of the root Ju first embodiment of the present invention is a remote health advisory system of the apparatus 110. ^ 3, the individual 110 comprises a statistical analysis unit 111, an individual correlation analysis unit 112 healthy individuals and individuals report generating unit 113 analyzing means. Next, each of the individual health analyzer unit 110 is described in detail.

Particular individual previous 140 individual statistical analysis unit 111 to the storage device and the health data currently input statistical analysis, the individual statistical analysis Students reflect the trend health l is the status of that particular individual previous and current input health data It includes a plurality of health data. Specifically, the individual statistical analysis of the statistics unit 111 health trends each individual item data for a particular input current and previous health data, such as health programs with the input data trends of time. : Trends health program data summarized health data, the trend directly characterize the health of the health project data can be directly reflected in the change of 1 to the current in the period of the health of individuals initially using the system, for example to He said rising blood pressure value in the range of above standard Huai, which often precedes the increasingly strict child hypertension. The trend of the results of the health of the individual is a summary of past health. Statistical analysis Statistical individual subject analysis unit 111 processing the current health data generated is transmitted to an individual report generator 113.

In the individual statistical analysis of the results of the statistical analysis of individual cells 111 resulting in health trends is characterized by the direct health trends like ^ that part of the health project data reflected. However, there are some health data, health data project is likely to affect the health of the data, such as living conditions, family status, lack of trends and health status data with these individual health programs from the statistical analysis results contact, so correlation analysis module 112 need fast behind particular individual and previous health data currently input for further processing.

Self correlation analyzing unit 112, configured to a particular individual memory device 140 in the previous and current input of health data and the health data items a particular individual health data and previous input current correlation algorithm executed in the previous and current derived particular individual healthy individual health data input arranged data items, and a combination of the health data items of the health data items are associated with the correlation result of individual health. Individual associations correlation analysis unit 112 analyzes can be performed to dig out the association between the health data items of health data. In the health data items, characterized health is directly associated health data items (e.g., blood pressure, etc.) or the like in combination with health health programs (e.g., hypertension, psychosis) is a direct and clear (e.g., higher blood pressure "blood pressure: 180/120" means suffering from hypertension). However, - in order to obtain 4 may produce effects on health, represents life status, family status health project data (such as "Today's drinking habits: Υ") associated with the health of blood pressure disease), you may need to dig out and the direct correlation between the data characterizing the health of the health data items or combinations of items health effects health health programs. For example, if the association between alcohol consumption and the higher blood pressure value is found, it is possible that a relationship between alcohol consumption and hypertension, it is enough to bear the result ί find out the cause of hypertension alcohol.

Above, "said today Drinking habits: Υ" and "hypertension" and shut ^ is an example of health data in the data associated with the health of the individual health programs. There is also associated with a combination of health of many of the health project data, for example, an individual association analysis unit 112 can produce the "Today Drinking habits: Υ": a combination of "Υ today whether to stay up late," the two health programs and data "blood pressure: 180/120" association, i.e. derived alcohol, stay associated with hypertension in combination of two events. There is also associated with certain health conditions arrangement indiscriminately term health data, e.g., individual analysis unit 112 with stars "exercise time: 30mins", "Today's drinking habits: Y", "Today's stay up if: Y" arrangement that the first movement, and then drink, then stay up all night, or before, drinking, and then exercise, stay up late, etc., and "blood pressure: 180/120" association, that can be drawn "exercise time: 30mins", "today's drinking habits : Y "," stay up if today: association γ "arrangement with hypertension. 'Find the correlation between the results of these objects are associated analysis unit 112 of the main function. After ffi same situation occurs several times, correlation analysis unit 112 that will be associated with such a strong event (such as "alcohol") in combination / arrangement or an event (such as "alcohol", "late at night") and higher blood pressure value to be sure, the user after experiencing a combination of these events, there will be a corresponding health. So next time when it experienced the same combination of events, the same symptoms may occur again. ,

Association Analysis of individual correlation analysis unit 112 performed by the algorithm is based on the correlation, which is a data mining algorithm, the transaction is composed of a plurality of transactions to find the concentration of each of the plurality of transactions comprises a firm algorithm for the association between the plurality of item sets. In association algorithm associated with the individual analysis unit 112 performed in the plurality of individual transactions with specific health data and previous input current transaction as a set, each particular individual health data and previous input current health data as in a transaction, using a collection of at least one health of the health data items for each particular individual health data current and previous input data composed of a plurality of items as a set of items set. In association algorithm associated with the individual analysis unit 112 performs, first set the required support and confidence threshold correlation algorithm, followed by the collection of data items to be exhaustive health health data (i.e., comprising one or more sets of health data items ) and calculating each set of data items each health health data (i.e., between the support and confidence of the set of items), the support between each of the calculated set of health data items, respectively, and support and confidence threshold and two or more confidence threshold, between the support and confidence of the two or more health data sets is greater than the threshold, then item association between two or more health data sets items, there will be associated with the last a health project data is saved as a collection of individuals associated with this result.

The following analysis of the correlation analysis unit associated with an individual embodiment of the first embodiment of the present invention in association algorithm 112 is described in detail based.

First introduced more important in the context of the algorithm several parameters, including the collection of transaction, transaction sets, items, item, item collection, support and confidence. In the first embodiment of the present invention, each time the user inputs, such as input said same period, all or part of the health data may be regarded as a transaction, e.g., the health data shown in FIG 2A and 2B may be as two different transactions, the transaction set consisting of a number of sets is called a transaction. Specifically, a plurality of health data items shown in FIG. 2A (or all of the health data items) is a transaction Tl, health items combining a plurality of data shown in (or all of the health data items) is a transaction FIG 2Β [tau] 2, transaction set D is a set composed of T1 and Τ2, D = {T1, T2}. Every health project data transaction is called a term, such as "Today's drinking habits: Y"; and if we consider the words of the sensitivity analysis for time entry is "Today's drinking habits: Y <2009-12-1>". Item set is a collection of health data items or more of a transaction, for example, item set Α =

( "Today's drinking habits: Υ <2009-12-1>"), item set Β = ( "blood pressure: 180/120 (mmHg) <2009-12-01>"). The correlation result is called an association between the set of items, for example, is an association between the item and the item set B set A, i.e., set B can be derived from the item set item A, i.e., ^ 4 S. Support means A and B, i.e., the probability that two itemsets appear in the transaction set D at the same time, the number of share transaction set comprises items in the transaction set D concentration percentage of all transaction number of transactions, and means confidence the probability that item set a appears transaction set D, itemsets B also occur, i.e. the transaction set. D, i.e. comprising a transaction item set B is the number ratio of the number of contained items set a transaction .

The association between the transaction may be immediate, or it may have a time interval. For example, a lot of drinking may lead to immediate headaches, but also cause long-term effects, such as two consecutive days of high blood pressure.

For immediate association, in which case health data in health data input time during the process is not taken into account. Can use existing transaction set correlation analysis, assuming transaction set D1:

{

{( "Today's drinking habits: Y"), ( "blood pressure: 180/120 (mmHg)")} - {( "Today's drinking habits: Y"), ( "blood pressure: 180/120 (mmHg))" },

{( "Today's drinking habits: Y"), ( "blood pressure: 180/120 (mmHg)") ( "Today's stay up if: Y").},

{( "Today Drinking habits: Y"), ( "today whether breakfast: Υ")},

{( "Today whether to stay up late: Υ"), ( "today whether breakfast: Υ"), ( "Today's drinking habits: Υ")}

}.

5 D1 transaction transaction set, there are three transaction comprising {( "Today's drinking habits";), ( "blood pressure: 180/120 (mmHg))"}, i.e., transaction 1, 2, 3, therefore, the results support this time was 60%; there are four transactions comprising {( "today's drinking habits")} in the 5 D3 transaction in transaction set, which contains the transaction {3 ( "blood pressure: 180/120 (mmHg ) ")}, then the result was 75% confidence level. Associated evaluation unit sets a certain threshold are support and confidence, when calculating correlations between different itemsets, if the calculated degree of support and confidence are more than respective threshold values, the correlation analysis unit that item there is a link between the sets. For example, if the support is provided at this time and the threshold value is 60% confidence level, the results can be drawn from the above calculation, itemset itemsets A and B are associated. In other words, from the associated itemset itemsets A and B can be derived alcohol can be immediately result in higher blood pressure value, in other words alcohol result in immediate (within this period records, produces effects) Hypertension disease.

However, some impact is not necessarily reflected in the day, such as the drinking can also cause stomach pains after two days, if you only use the transaction record set the day, such an association can not be be reflected. In order to reflect this difference hidden in time relationship, individual correlation analysis unit 112 needs to re-set the original transaction combination, the combination of transaction specific intervals into a new time-sensitive transactions, and these newly synthesized the timeliness of the transaction will constitute a new transaction sets timeliness, timeliness refers to the process is to consider the health data input time health data. In this case, selected from a particular individual, and previous health data current inputted in pairs health data combination, and in combination of the related health data as transaction algorithm, wherein, in pairs selected health data of the health data input time difference to a predetermined value; transaction data includes all health items in pairs in a selected health data; and the selected pair of health data item in at least one health a set of data as a plurality of items composed of a set of items set. Herein refers to the combination of the data will be combined in pairs in a manner selected health all health data items which correspond to the same items, but the health data items corresponding to different health health data input time difference from each other, e.g., "today whether drinking: Y <2009-12-1> "and" today's drinking habits: Υ <2009-12-3> "in the affairs of health as a combination of data processing as two different projects. For example, the transaction set D rfa original recorded daily, 7 / = {T1, T2 , ..., Tn}, where Ti represents the recording from the recording starts its i-th day. If we are interested in the associated timeliness of two days, then the transaction is a set of time intervals two days (i.e., the next day) and generates a two days interval (every other day) transaction set D 2.,. = {T1 + T3, T2 + T4, T3 + T5, Tn-2 + Tn} 0 configuration transaction timeliness in item will show time. D 2i ^ such as:

{ " '·:.. ·, ·:. ::

{( "Today's drinking habits: Y <2009-12-1>"), ( "blood pressure: 180/120 (mmHg) <2009-12-03>")}, {( "now latch drinking habits: Y < 2009-12-2> "), (" blood pressure: 180/120 (mmHg) <2009-12-04> ")}, {(" now said drinking habits: Y <2009-12-3> "), ( "blood pressure: 180/120 (mmHg) <2009- 12-05>"), "whether or not to stay up late today: Y <2009-12-3>"},

{( "Today's drinking habits: Υ <2009-12-4>"), ( "Breakfast whether today: Υ <2009-12-6>")}, {( "Today's stay up if: Υ <2009-12-5 > "), (" today if breakfast: Υ <2009-12-7> "), (" now reads drinking habits: Υ <2009-12-5> ")}

} Ο,

"Today Drinking habits: Υ <2009-ί2-1>" records from December 1, 2009, and "blood pressure: 180/120 (mmHg) <2009-12-03>" from the record after two days.

5 D3 transaction transaction set, there are three transaction contains { "Today's drinking habits <T>", "blood pressure: 180/120 (mmHg) <Τ + 2>"}, wherein, Τ a specific date, Τ + 2 is after the specific date two days, i.e., transaction 1, 2, 3, therefore, the results support this time was 60%; there are four transactions contained in the five transactions in transaction set {D3 ( "today's drinking habits ')} ,. transaction contains {3 wherein "blood pressure: 180/120 (mmHg"}, the results at this time was 75% confidence level associated evaluation unit sets a certain threshold for the confidence and support, respectively. when calculating correlations between different itemsets, if the calculated degree of affection and a set of support exceeded the respective threshold, the correlation analysis unit that is associated between itemsets. for example, if at this time threshold setting support and confidence is 60%, the results can be drawn from the above calculation, item set a and set B associated item. in other words, the set of associated items a and B can be derived itemsets drinking can lead to higher blood pressure value after two days, two days will result in other words, after drinking Gao blood disease. in Case, it is possible to find out that is associated with immediate relevance and timeliness of the find, this embodiment is only interested in the timeliness of the association, which has shut Zhen health project data between different health data input time. Also , where the next day (day apart) of example only, such a case is also equally applicable to the spacer angle days, like every other week.

Individual association analysis unit 112 will be exhaustive something set D1 in all entries set to continue to look for other item sets are related, in the above example, only one item (health project data) item sets A or B, the association analysis mining is an association between the individual health data item, when the item set a or B are arranged in a plurality of items or a combination of a plurality of items, the elapsed treatment similar to the above can be obtained by a combination of the health data items / the association between health status and health programs aligned data. When the association can no longer find in the transaction set D1, the association analysis is terminated. Correlation analysis can be implemented in a variety of specific languages, e.g. Aprior algorithm, as is well known to those skilled in art, the present invention will be omitted for the introduction of the specific implementation of the algorithm Aprior.

Correlation analyzer self correlation result generated 112 can exhibit health data in a single healthy item on personal life status, family status and other data, a combination of health program data / arranged directly characterizing health health programs or health project portfolio association between, thereby showing off ¾ health data results rather healthy individual data items, a combination of the health data items, the data items are arranged in the health associated with health. Correlation result is transmitted to the individual subject report generator 113 for further processing.

Subsequently, the individual may report generator 113 based on a particular user's previous reading from the storage device 140 and the input current health data, the statistical analysis of individual and the individual correlation results generates a healthy individual analysis report in a predetermined format. In other words, the individual ¾ Kang analysis report generation unit 113 individuals included in the generated current and previous specific user input health data, statistical analysis of the results of the individual and the individual correlation results.

3. Group health analysis means 120

4 illustrates a remote server in a group as a group 10 wherein the analysis means 120 example of analysis apparatus next with reference to the health. Group health analysis means 120 generates a plurality of groups based on all the previous individual memory storage device 140 and health data currently input, Statistical Analysis and Results The results with the associated group corresponding to the plurality of groups, specifically, groups group health analysis means 120 reads all individual calendar from the storage means and the current input 140 ^ health data and generate a plurality of cluster groups thereof, and no multi-group statistical processing, correlation analysis, respectively, to generate group Groups the statistical analysis and the results of group association, and all individuals based on health data and the previous input current, a plurality of groups, group and groups associated with the statistical analysis result generating analysis report as examples of the group feature health analysis report. FIG 4 is a block diagram showing an exemplary configuration of a remote health apparatus 120 according to a first embodiment Consultation embodiment of the present invention, the group 1 ^ health analysis. Group health analysis apparatus 120 includes a clustering unit 121, the statistical analysis unit group 122, group association analysis unit 123 and report generator 124 groups.

First, the clustering unit 121 is based on a combination of all of the health data items or single items of health, the health of each group pair analysis apparatus 120 storage device 140 all the current input and previous individual health data clustering algorithm execution object user health data clustered into a plurality of groups (a first plurality of groups). Corresponding to the plurality of health data in a first group generated by the group based on the individual health programs generated, which is present in the same group is the same as the single health wellness programs or data item in the same data range within; health data corresponding to the plurality of health of a group of items based on the generated, which is present in the same group in the plurality of data items of the plurality of health wellness programs are the same or are within the same range of data. Described below to a specific example of the process of clustering the plurality of health data in accordance with each combination of individual items or health health programs. For example, there are four individual A, B, C, D, has their health programs and data items are A: (occupation = 1, height = 160, revenue = 1000), B: (2 occupation = Height = 180, revenue = 1300), C: (= 1 occupation, height = 170, revenue = 2000), D:

(Occupation = 2, height = 185, income = 3000). If the health programs based on "occupation", clusters of four individuals, the same occupation, i.e., as a health ¾ program data represent the same individual is the specific occupation code into a group, then A and C are poly in one group 1, B, D in the group 2. When the cluster-based health programs "height", the group A in Gamma] (height <165) and, B, C, and D in a group 2 '(height> 165). When the cluster-based occupational and height, A in the group 1 "(= 1 occupation, height <165),, C in the group 2 '

In (occupation = 2, height is greater than 165), D in the group 3 "(occupation = 2, height> 165) In the time-based clustering height and earnings, A, C Gamma] in the group" (height <165 <in 1000), B, D in the group 2, "(height> 165 income, income> 1000). according to the same single group or a plurality of items of the same health items combining the generated sound it is referred to as the same class combination group, so clustering unit 121 generates a plurality of types of the second group, i.e., for example, when health data., the clustering algorithm. the number of individual items and a plurality of health health Initiative is the sum of the number of N (second plurality), it generates the N types of groups, the total number of these groups are the N types of M (first plurality).

In the clustering process, for many obvious standard is of great significance, the system can be preset clustering criteria, such as the above-mentioned career clusters, as well as gender clustering. However, for complex standards, labor standards preset completely incompetent, for example, what kind of income by more than standard clustering considered science? Age, height level how clustering is reasonable? Let alone for a combination of two or more clusters of projects. At this point, we need to take advantage of clustering algorithms. Clustering algorithms need to construct a dissimilarity matrix, and matrix data from each user's health data, health data re-clustering the data matrix and to dissimilarity matrix. The following details the clustering unit 121 clustering algorithm executed in the health data. Clustering algorithms require different individual health programs together with all of their health data in the data represented by a data matrix. For example,. N-users, each user having a healthy P project data value such as height, weight value, the occupation code, etc. Such a data structure can be represented as a matrix nXp, as follows:

Clustering algorithm also requires a dissimilarity matrix, [eta] which includes all pairs of adjacent users, usually η Χη dissimilarity matrix represented as follows:

Which is a measure of the difference or dissimilarity between the individual and the individual's health data. Typically, i (i, is a non-negative value, the more similar or more similar individuals and health data, the closer to 0, the greater the difference between the individual and the health data, the greater its value. At the same time, d 'DH and / (/) = 0. dissimilarity may be used to calculate various types of data, including interval-scaled variable (e.g., height, weight, etc.), symmetric and asymmetric binary variables (e.g., gender, or whether drinking (Y or N)), categorical variables (e.g., or occupational), ordinal variables (e.g., posts, etc.) and a variable scaling ratio (linear scale, such as scale index) and the like, or a combination of these variables'.

c (,> specific calculation methods and procedures are well known to the skilled person and are therefore not repeated here. performs clustering algorithm may take many specific calculation methods. For example, in the division method used in the present embodiment embodiment (Partitioning method) in the k-means algorithm. in the process division method, first, to select individuals from a given k n individuals randomly, k individuals each individual health data (health items containing p) represents a mean initial clusters, ··· for each of the remaining individual U = nk), each cluster according to its distance from the mean, namely: for k clusters average, d (, x, j = (1, 2, ... K), it is assigned to the minimum clusters after recalculating the mean of each cluster as familiar to the person skilled in the art, the mean can be defined as follows: three-dimensional data set, a data cluster has two healthy : health data χ = (^, ...) and health data Y =

. (, 2, · '· ν) Mean [zeta] is: Ζ = (, · · · ), wherein Zl = (Xl + y,) / 2, z 2 = (x 2 + y 2) / 2, z n =

(x n + y n) / 2. The new cluster mean, every object will be re-assigned to the cluster, the cluster continues to recalculate the mean. Repeat these steps until convergence criterion function. Typically, a squared error criterion. In other words, for each object in the cluster, to find the object from the center of the square and its cluster. The guidelines attempt to minimize the E to the cluster as compact as possible and independent.

In addition to these methods, the skilled person is well-known hierarchical method (hierarchical method), the density method (density-based method) based on a grid approach (grid-based method) based, model-based methods (grid-based method )and many more. Various methods of performing clustering algorithms listed above are illustrative, as is well known to those skilled knowledge, not going to be described in detail herein. The method of performing clustering algorithm used in the present invention is not limited to the above, further include any other suitable method in the art to understand the art.

A plurality of groups after clustering unit 121 clusters generated is transmitted to the back of the device.

Next, the group statistics analyzing unit 122 further analyzes the plurality of the first group. Statistical analysis of the first group of the plurality of unit groups each of a group of health data for statistical analysis, generating a first display health clustering unit 121 generates a plurality of groups in each of a group of statistical analysis of the results of the group. Specifically, the individual statistical analysis unit 110 is similar to the processing means analyzing an individual's health, group 1 1 1 Statistical analysis unit 122 by a first plurality of groups each group in each of the health data All health project data for statistical analysis, especially health project data can be directly characterize the health of statistical analysis, statistical analysis of the results obtained can exhibit the typical health of the group. For example, when the statistics of the blood pressure value (i.e., the higher blood pressure value) exceeds the standard value of the health data group included in the frequency of occurrence exceeds a certain threshold, then the group can be derived "higher blood pressure value." frequent, that hypertension is typical of the group's state of health. Group is transmitted to the statistical analysis group report generation means 124. 'Next, the analysis unit 123 groups associated with a first group of the plurality of correlation analysis process. Groups associated with the first analysis unit Health Project clustering unit 121 generates a plurality of groups in each of a group of health data and perform data associated with the same individual health algorithm 110 executed by the analysis means, to give the result of the combination groups associated health programs or health data items in the first plurality of data groups each of a group of health data respectively associated with the M state of health. Association Analysis group association analysis unit 123 is performed based on the same principle of self-association analysis unit 110 and analyzing means 112 healthy individuals, but is within the range of each group comprising a group of healthy correlation analysis to identify the data correlation results, specifically, all of the health data sets within the group as a transaction, and each transaction as health data, to associate an individual analysis unit 110 and the individual health analyzer 112 same correlation analysis to obtain the correlation result, wherein the correlation algorithm is a centralized algorithm to identify the association between the plurality of item sets each of the plurality of transaction transaction contains multiple transactions in a transaction thereof. In association algorithm, all of the health data using a first plurality of groups each of a group included as a transaction set, a first plurality of groups each of a group of health data as a plurality of each transaction in a transaction, using a set of health data by each of a first plurality of groups each of a group of at least one of the health data item as a plurality of items composed of a set of items set.

Also note that, with a different group analysis is focused on the individual transaction record is an individual, and group analysis, the records may be derived from different individuals, so the build process item sets, the group association analysis unit 123 according to the source needs to be recorded screening to ensure that only healthy and projects from the transaction data can be combined in the same individual integrated transaction set and items used in the analysis, to avoid confusion between the different individual records. SP: A and B belong to the same individual in a group, A a particular drink, will only lead to headaches A, and B on the same day if it happens stomach pains, and A drink should be nothing to do. In this case, first of all the previous individual health data and the current input of health data belonging to the same individual comprising two or more health data, from all the individuals belonging to the health data and previous health data currently input to into the same individual mode selection for the combination of health data, and a combination of the health data transaction as an associated algorithm, wherein the difference between the input time of the health data in pairs selected health data to a predetermined value; transaction comprises in pairs selected health data items in all the health data;. and at least one pair of the selected item healthy health data in a data set composed of a plurality of pages as a set items set I. Combination here refers to all of the health data items will be combined pairwise manner selected health data, health programs which correspond to the same items, but the health data corresponding to different health data input time difference from each other.

For this case another treatment is: only records the individual identification codes of the individual services and items belong, so their group association analysis unit 123 records rejected association between unrelated individuals A and B of the subject, because its associated contingency it is difficult to reach a threshold analysis, but can not constitute the association rules.

Finally, group association analysis unit 123 analyzes the correlation results of each group as a group to the correlation result transmission group report generation means 124.

'Individuals and individual health report analyzer 110 113 class generating unit, the group generation unit .124 She reported based on a first plurality of groups obtained category unit 121, the statistical analysis unit 122 in the group generated in the group statistical analysis results are set in a predetermined format and generating a correlation result in a group associated with the group analysis unit 123 generates a group health analysis. In other words, the group association analysis unit 123 generates group health analysis package ¾ of a group, the group and the groups associated with the statistical analysis results. From the above description, the statistical analysis unit analyzes the individual apparatus in healthy individuals 110,111 and individual analysis unit 112 associated with the group health group Statistical analysis unit 122 and the group association analysis unit 120 analyzes the device 123 may each share a statistical correlation analysis unit and the analysis unit.

4, a comprehensive health analysis device 130

Next, with reference to FIG. 5 explained in detail, general health analyzer 130 remote health advice system of the first embodiment of the present invention 1. Analysis of Health data group after the beginning of the Healthy ho processed device 120, a remote health advisory system completed; healthy individual analyzer 10 ^ 1. To obtain more accurate and comprehensive health advisory report, the need for health data group and individual health analysis devices for further comprehensive analysis of more than 120 groups drawn. To understand that there will be some correspondence between the plurality of groups of individuals and health data generated by the group health analysis means 120, in other words, the health data corresponding to the individual to be classified according to its own situation into each group. Scientifically correctly classify individuals into various groups will provide individual health records of good health background and reference data, such alignments can understand the current situation for personal health, the health condition of the ^ trend of the future forecast to provide more information. For example, a 30-year-old consultant, their health trend 10 years later, we can present to the overall health in the form of 40-year-old consultant and as a comparison reference. If it finds the age of 40, his health could be seriously affected, and to find possible causes, combined with the guidance of their health related analyzes the results, and third-party experts, you can take precautions ahead of time, or improve living habits in order to avoid possible health risks.

Examples of general health analyzer means analysis according to the first embodiment integrated feature of the present invention as the embodiment 130, general health analyzer 130 depending upon the particular individual stored previous storage means 140 and current input Qin of health groups and health data a plurality of groups (a first plurality of groups) generated by the analyzer 120 generates a plurality of groups (the second group) and the second relates to the optimization of the plurality of groups corresponding to the specific previous one input health data results of predictive analysis, specifically, a comprehensive analysis of a plurality of first group 130 Laid Zhongluo previous individual read from the storage device 140 and health data and the current input means 120 generates a group of the healthy perform classification, - prediction process generates a particular individual health data corresponding to the previous input group and the second plurality of results of predictive analysis relates to optimization of the second plurality of groups, based on the second plurality and the second plurality of groups and ^ analysis of results of predictive analysis to optimize the group generate a comprehensive health.

FIG 5 is a block diagram showing an exemplary configuration of the analysis device 130 according to the comprehensive health remote health advisory system of the first embodiment of the present invention 1. Comprehensive health analysis device 130 includes a similarity calculating unit 131, classification 132, the prediction arithmetic unit 133, the integrated optimization and synthesis unit 134 output unit 135 unit.

Calculating a first plurality of groups between a similarity calculation unit 131 based on the similarity of a particular individual algorithm stored in the previous and current input means 140 Po health groups and health data analyzing means 120 generates in each group to obtain reflect the degree of similarity between the previous and current input the particular individual health data and each of the plurality of similarity group. Similarity calculating unit 131 calculates the similarity clustering process the same apparatus 120 described in the dissimilarity of the Healthy c in the group ( "j calculation process, the similarity calculation unit 131 calculates the particular individual's previous and current input health data and the center point of each group, or the degree of similarity between the above-described through cluster mean. similarity calculation unit 131 a plurality of similarities obtained are saved for subsequent processing used as the art techniques well known techniques, will be omitted specific explanation of the process of similarity calculation.

Next, the classification processing unit 132 to classify the particular individual input current and previous health data and a plurality of the above groups based on the similarity calculated by the similarity calculating unit 131 to select a particular individual from a plurality of groups previous and current input health data corresponding to a plurality of groups, i.e., a second plurality of groups. It referred to herein, "corresponding to" that corresponds to the particular subject "is classified into" into a plurality of groups, i.e. specific health data ^ Most individuals are included in their health data included in the group are very similar. Specifically, the categorization process is in a group of healthy analyzing means 120 generates N (N <M) number (a second plurality of kinds) of the M (first plurality) group, collation unit 132, respectively from the N types of groups each of one species of the group, Yu selected when health data input group with the highest similarity particular individual calendar ^. That is, the clustering section 132 selects M groups of N out of N types corresponding to the specific subject of the current input and previous health data group, i.e., a second plurality of groups. For example, after a given individual of previous and current health data input means are classified, the individual may be classified as the group X health analysis means 120 generates the M groups (a first group of the plurality) of the N groups, M groups in the rest of the (MN) th group is removed, to give the final N groups, i.e., a plurality of groups corresponding to a particular member of the current input and previous health data (second multiple groups). A second group of a plurality of collation unit 132 are obtained and stored for use to process the next ho.

Next, the prediction arithmetic unit 133 a specific input current and previous health data and perform a second group of a plurality of prediction algorithms to obtain a second plurality of the plurality of predictive prediction group health change correspondence analysis As a result, that is, the second plurality of groups each group has a phase of predictive analysis results for ^ ^ health conditions. Process analysis unit 132 in a prediction for predicting the value of the continuous data object prediction bin method, a particular individual input current and previous health data and parameters as part of the decision nodes on the predicted path, whereby the can the results of each of the prediction groups to obtain a second plurality of groups.

Prediction algorithm needs to generate predictor. 6A is a schematic view of a process of generating analysis Kenneth exemplary predictor of risk for hypertension consultants example, FIG. 6B is a risk for hypertension for analysis of consultants illustrative examples of the page to the detector, FIG. 6C are exemplary examples of a plurality of predictive analyzes risk of hypertension different groups, wherein the health data items and health programs provided merely illustrative. Prediction algorithm implementation process, most groups use predictor of health data, namely the use of all belong to this group and occupation of this post, for example, a group of health "consultant" professional job "consulting" data, and specific health conditions, such as hypertension, get predictable results, such as high blood pressure disease risk Gao, medium and low.

For example, X may be assigned to different user groups of eight, B, C-, if a particular health condition we are users of α (eg, hypertension) are interested, we can construct with each group predictor corresponding group, such group Α - α predictor group Β - α predictor, group C - α predictor, and the like. Each predictor predictor will produce a corresponding α, as a result risk prediction I, II risk prediction result, risk prediction results III. Depending on the use of samples, the prediction algorithm to predict the results will be different. The predictor constructor kind of known methods, for discrete data, including decision tree classification, naive Bayes, Bayesian belief networks, and support vector machines. For continuous data, using linear and non-linear regression (Regression) method. In this invention, various algorithms can better complete the task, in order to facilitate expression, we use a decision tree is described here.

Decision tree induction is from a training class mark (particular health condition) tuples learning tree. A decision tree is a tree structure similar to the flow chart in which each node represents a health project (such as post ^^ income, lifestyle, etc.) on the test, each branch represents a test output, and each leaf node to store a class label (forecast results for a specific medical condition). Topmost node of the tree is the root node.

Most decision tree induction algorithms follow a top-down approach, beginning from the tree structure (the predicted results for a specific medical condition) and training tuples and their associated class labels, with the construction of the tree, the tree training recursively divided into smaller branches. The system selects all items generating a healthy test node in a particular order, until a branch in this node with the same class label (specific health predictions).

Prediction analysis unit 132 will first According to a particular "measure attribute selection" as the splitting criterion, it is determined in the decision tree structure from top to bottom. Such as attribute information gain metric, i.e. preferred splitting attribute has attribute information gain as the highest node N, this property is the amount of information needed to classify the results into a tuple in a minimum, and to reflect a minimum of these partitions and randomness "impure."

Gain information can be derived by the following equation:

Mo (D) = -f j Pi \ og 2 (Pi)

/ = 1

Wherein, A is the probability of D belonging to any of the principles predictors C, and with | C ,,. | / | Z) | estimate. If you choose healthy items A, such as "marriage" as selected health programs, in accordance with the decision tree of all its property values ​​possible health project data (project data that is healthy for all possible values) Αμ ,,, ^}, or {marriage "married", "married", "divorce"} are classified into three branches of the decision tree.

If all property values ​​A health project data as a continuous value, such as income, you must first determine the "best" split point A, where the split point is on the threshold of A. A first refers to the need to sort in ascending, typically, the average value between two points can be seen as a possible split points. Thus, A is a given value of V, v-1 to be calculated possible division is calculated for each possible Info A split point (), select the point having the highest information gain as the best split point, which corresponds to information gain, as the gain information of this property. If this attribute is selected as the continuous division criteria, the best tree "split point" (spl, it_point) as the threshold value, it is divided into two branches according to: A <= split_point, A> split_point.

In addition to information gain, as well as gain ratio and Gini index and other indicators can be used to determine the value of the attribute data of selected health programs.

When a particular predictors, such as a decision tree has been constructed and tested and optimized net. Other data may be recorded in accordance with the analysis result of the decision tree from top to bottom until it found the desired result of the prediction.

Described below to predict one example of the obtained results. To Figure 2A and 2B data shown, for example, 35-year-old foreign consulting firm consultant Zhang, 35 years old belong to the group, the group consultant, foreign-funded groups, married with 2 'children. For each group will be given a certain degree of Zhang belongs predictive analysis results for specific health conditions. For example, for the prediction of hypertension Mr. Zhang, the first specific health data through predictive algorithms hypertension consulting industry consultants to obtain specific risk of hypertension, for example, 75%; again to Zhang health data through prediction algorithm with children group, their risk of hypertension was 50%; then, its data through predictive algorithms regular exercise group, the risk of hypertension was 25%. Respectively, and optimization of the group after the above process the results of analysis of multiple groups of health information contained in correspondence reflects more predictive analysis result of changes in health status, such as, respectively, a group of consultants with children three hypertension risk groups, regular exercise group one to one, that is, results of predictive analysis is respectively 75%, 50% and 25%, these predictive analysis results are saved and sent to further processing. However, whereby, in order to get a clearer an individual prediction results, it is necessary to use a plurality of the above-described 134 pairs of results of predictive analysis unit to optimize the integrated optimization.

Integrated optimization unit 134 according to the above-described plurality of similarity obtained by the similarity calculating unit 131, a plurality of predictive analysis result to generate a weight distribution to optimize the results of predictive analysis. By referring to a particular individual similarity between the current input and previous health data and a second plurality of groups corresponding to a particular individual and previous health data currently input, you can draw a particular individual input current and previous health data for different propensity prediction algorithm. Based on the similarity, the integrated optimization unit 134 analyzes the plurality of predictive information for health weight distribution, for example, be a weighted average, weighted according to the size of the different similarity decisions, to obtain the final prediction optimization analysis information. Zhang to the example above, the consultant Zhang in the clustering tendency is moderate (assumed to be 0.3), families with children and low tendency to (assumed to be 0.1), while in regular fitness clustering Higher is the (assumed to be 0.6), it may be the result of a weighted average of 75% + 0.3 * 0.1 * 0.6 * 50% + 25% = 45%. The above conclusions can be Zhang suffering from hypertension risk was 45%, which is the result of the analysis of individual and group analysis of the resulting combined, the result is relatively higher accuracy, which is a state of health the accurate and reliable results predicted.

Weighted average calculation above is a simple illustrative example. Which is based on the similarity between the previous and current input the particular individual health data and a second plurality of groups, each group of the prediction results obtained weighted results. Further, more complicated formula for calculating weight distribution may be, for example:

E (x) =. F + (5) where E (x) is the result of the prediction of individual x; n is an individual, and belonging to the n groups; parameter with respect to the i-th group; X is an individual with respect to the similarity of the group i; =: is a predictor of group i; i group with respect to the reference power of similarity; and is relative: the power to groups i predictor parameters.

However, the calculation is a weighted Λ relatively simple initial approach. With the increase of data, the formula will be continuously optimized in order to more accurately fit with the actual situation. , Such as adding a parameter in a formula, or the use of different formulas like a power of different items. Formula can also use a median threshold, and the like to fit the data.

Finally, a comprehensive report of the analysis result output unit 135 generates a predetermined format W analysis predicted overall health and to a second group of a second plurality of the plurality of groups based on the optimization, i.e., comprehensive report on the output unit 135 the generated general health a second analysis comprises a plurality of groups and a second plurality of groups involved optimization of results of predictive analysis.

Thus, remote health advisory system 1 through the above-described apparatus 110 analyzes an individual's health, the health analysis apparatus group 120 and general health analyzer 130 for health data processing, analysis generates individual health, health group analysis and synthesis report.

Finally, the output device 12b receives the output of said individual's health and analysis report in a predetermined format, a group of health and general health analysis analysis.

From the above description of the remote health advisory system wherein a remote consultation system as an example, the features can be summarized as a remote consultation method for remote health advice health 'advice example of a method, the method comprising: input processing, collecting input health data, health data including individual projects and individual identifier data item identifier, a number of health programs and health programs and the corresponding health project data;

Storage process, depending on the individual health data stored identifier data items in the health data;

Individual health analysis, system Review analysis and individual correlation results for a particular subject of previous and current input health data generation individuals According and ¾ to a particular individual previous and health data currently input, the individual statistical analysis and individual correlation results generates individual health analysis report, .

Individual health analysis process further includes:

Statistical analysis of the individual steps of the previous particular individual health data and statistical analysis of the current input to generate individual statistics reflect the trend analysis of health status results;

Self correlation analyzing step of performing a single correlation algorithm obtained health project data, health data items and combinations of data items a particular individual health previous and current health data inputted to the health data items a particular individual input current and previous health data individual alignment results are associated with the associated health, wherein the correlation algorithm is to find the correlation between the plurality of focused item sets each of the plurality of transaction in the transaction by a transaction contains multiple transactions composed algorithm, a particular individual previous and current input of health data as a transaction set, each particular individual health data and previous input current health data in the plurality of transaction as a transaction, and is used by a particular individual input current and previous health data each of the health data items in at least one health data set consisting of a plurality of items as a set of items set, when the case of time sensitive health data, the correlation algorithm, the particular individual and previous health data from the current input as a selection of health data into transactions, Health data input time difference between the two pairs of health data in the selected health data is a predetermined value, the transaction comprising all of the health data items selected pairs of health data; and selected in pairs set of health data in at least one health data item composed of a plurality of items as a set of items set; and

Individual report generation steps, the individual results of statistical analysis and generate results associated with individual health individual analysis reports based on specific individual input of all previous and current health data. '

Group health analysis processing to generate a first plurality of groups based on all the previous individual health data and the current input, statistical analysis group and groups associated with the first plurality of results corresponding to the group, based on the first plurality group, group and groups associated with the statistical analysis results generated group health analysis report,

Group health analysis process further includes:

Clustering step, respectively, and a plurality of items based on the individual health wellness programs stored on various individuals for all processing of the current input and previous health data clustering algorithm is executed to generate a first plurality of groups, a first plurality of groups, health data corresponding to a group of items based on the individual health generated, which is present in the same or in the same group within the same range of data in a single item health health data item,

A plurality of groups based on the generated health programs, which are present in the same group corresponding to the plurality of health data in a plurality of data items health wellness programs are the same or are within a data range ^; group statistical analysis step, the health data of each first group of the plurality of groups ΐ statistical analysis, a first group to generate a display of a plurality of groups each group of health of the statistical analysis; group set associative analysis step, a first plurality of health data items in each group a group of health data association algorithm executed, obtain a first plurality of groups each of a group of health data health the results are associated with groups associated with health data item or combination of items of health data, the correlation algorithm, a particular individual health data and previous input current transaction set as a specific individual and previous health data in the current input each health garment as a transaction number of the plurality of transaction, each used by a particular individual and the previous and current input health data At least one set of health data item health data composed of a plurality of entries set of entries set; inch pleiotropic case having health data, i.e., the health data further includes health data input time, input current and previous particular individual the health data includes two or more health data, selected from a particular individual, and previous health data current inputted in pairs health data combination, and in combination of the related health data as transaction algorithm, wherein, in order to health data paired manner selected health data input time difference to a predetermined value, the transaction comprising all of the health data items selected in pairs of Granville data; and

A set of data into the at least one characteristic item selected in pairs of the plurality of feature data items as a set of items set; and

Ho group report generation step, based on a first plurality of groups, group count and group association analysis result generating analysis healthy group.

General health analysis process, according to the particular individual health analysis and health data and the previous group of current input processing for generating a first group of a plurality of previous inputs to generate a particular individual health data corresponding to the second group and to a plurality of second results of predictive analysis to optimize multiple groups, and generates reports based on a comprehensive health analysis and optimization of the second plurality of groups of predictive analytics unravel fruit.

Comprehensive health analysis process further includes:

Similarity calculating step, calculated in a particular individual health data and previous input current health analysis with a group of a plurality of take off clothes similar similarity between the first process of generating a plurality of groups in each group, respectively; grouping step, the plurality of similarity based on a particular individual input current and previous health data and perform a first plurality of classified groups, an input to select a particular individual from a previous and a current of a first plurality of groups a second plurality of data corresponding to the healthy group;.

A plurality of prediction algorithms step performing prediction algorithm Laid ¾ individual input current and previous health data and to derive a second plurality of groups correspond to the second group of the plurality of predicted changes in the health status of the analysis result; integrated optimization step, redistribution optimized prediction to generate a plurality of analysis result of the similarity of a plurality of results of predictive analysis for the right;

Comprehensive report generation step of generating a comprehensive health analysis reports based on the results of the second plurality of predictive analytics and optimization group.

Output processing, and receives the output of individual health analysis report in a predetermined format, group health analysis and comprehensive health analysis report. Λ, (first modification).

Referring now to Figure 7 describes a first modification of the first embodiment. FIG. 7 is a block diagram of an exemplary remote health advice system of the first modification of the first embodiment of the present invention according to la. In the following description of FIGS. 7 and remote health advisory system la a first modification, the same portions of the first embodiment will use the same reference numerals, and description thereof will be omitted. In addition, for the convenience of explanation, the following in order to apply for individual health counseling as an example of a remote health advisory system la will be described. The remote health advisory system la in the application of the health status of the population in consultation only with the same change in the first embodiment it.

In the process of using a remote health advisory system la, the individual can by simply entering text or voice information in the form input health data added content free to enter, which includes some of the self-perceived health status, the user's feelings about living data habits, work status and a variety of activities vary and may be difficult to rely on a standard questionnaire obtained, or an individual thought to affect the health standard questionnaire but temporarily not preclude housed.

In such a scenario, the remote health advisory system must la data identification means 160, so that after a certain processing health data between the memory storage device 140. Identifying data from the input means 160 consisting of individual content items identified additional health and the health of the corresponding data item, and additional health programs and data into the corresponding item health health data, health programs in additional health data to individual freedom enter the mode of existence of the project.

Identification data processing device 160 consisting of individual input contents as follows. First, the information input device 160 to identify the data required to identify and to capture. In general, important information including the time, the user's experience, health, and other activities. Initial stage, data identifying means 160 lists the corresponding data format, such as time, frequency, and associated keyword synonyms, antonyms dictionary for capturing information. With the increase of data, specific information, such as high frequency of occurrence of the keyword will be recording data recognition means 160, and added to the dictionary. With the increased use of individual users will follow rich personal dictionary, data identifying means 160 generates a dictionary for each user on the basis of the initial dictionary. Enter the contents of individual freedom can directly reflect the content of some health, some can indicate mood, living conditions, living habits, there are other exercise can help data on health status. Here is an example of some keyword categories, examples of which include health, stomach pain, bloating, depression, insomnia, pain, fatigue, full of energy, loss of appetite and so on; examples include mood and feelings, happy, happy, unhappy good mood, depressed, unhappy, sad, frustrated, restless, ill at ease, calm, agitated, uncomfortable ... and so on; examples of habits and conditions, including, just wages, friends, carnival, overnight, overtime; represents a frequency examples include, X times, X bottles, X times, all day, a week X number of times ... and so on.

Keywords will be recognized as a valid health data of individual freedom typing join the transaction record, as the object of further analysis. Further, according to the well-known natural language processing, and the semantic analysis can be logically simple sentence input by the user, it has to obtain more detailed information.

For example, if the input content includes individual freedom "to work overtime today for a long time, especially fatigue, stomach starts its feel pain." For this statement, we can extract the corresponding health data. For example, the foregoing "overtime" may be understood as the reason, feeling of fatigue, stomach pains and the results can be considered. This phrase can be used as a transaction «" long overtime "), (" fatigue ":), (" stomach ")} recorded. Data identifying means 160 to establish additional health programs based on the content data of the identified different, and identified as the corresponding receiving health data item, as the example, input 1 can be drawn freely = overtime long, consisting of input 2 = fatigue, stomach free input = 3, shown in Figure 2B. Wherein the free entry of 1 indicates living conditions, free to enter 2 and 3 are characterized by free input state of health.

Health data measuring instrument contains items, consisting of a standard questionnaire items and individual items are then input to 130 as in the first embodiment is processed in Example 110 healthy individuals, the group health analysis means 120 and analyzing means analyzing apparatus overall health. Health care input data items to enrich health health data to the individual data items consisting of individual input generated content, such that a more comprehensive remote health advice based on the analysis system of the first embodiment of the present invention 1.

In addition, including the case of individual freedom input data in health programs in health data, health programs due to individual freedom input data contains a large number of symbols entity (such as key words and phrases) and other complex objects (such as documents), therefore, in the group clustering algorithm execution means 120 for analyzing the health of the distance between the measuring complex object, typically to introduce a non-metric similarity function. For example, to compare two vectors X and y, there are several ways to define such a similarity function s (x, which includes a cosine similarity metric function is defined as: Wherein, c 'is the transpose vector X, vector X is the number of the Euclidean norm. S In essence, is the cosine of the angle between the X and y to. - similar to the above-described measure a variant thereof can be used alternatively, i.e. Tanimoto coefficient, Tanimoto or from s (x, y) = - ^ - (4)

x · χ + y - y - x ■ y

(Second Modification)

Remote health advisory system such as data processing function is more complete, in the first embodiment, the remote health advisory system according to a second modification of the first embodiment further comprising a further lb some other device components. Identical to the first embodiment with the means and unit in the first embodiment, the same reference numerals, description thereof will be omitted. Referring now to Figures 7 and 8 illustrate a remote health advisory system according to a second modification of the first embodiment of the present invention lb. FIG 8 is a detailed system design remote health advisory system lb a second modification of the first embodiment of the present invention. FIG 9 is a diagram illustrating the connection party to provide remote health advisory system according to a second modification of the first embodiment of the present invention with an individual lb health advisory service usage. 9, individual, i.e., the user mobile device, a desktop device or a home monitoring station via the Internet or a commercial wireless network service provider located remote health advisory system at the side of the remote server to exchange data.

a. The input device and the output device 12a 12a

Output means 12a and input means 12b comprises a plurality of instruments for measuring health data, for simplicity meter, in FIG. 8 exemplarily listed instrument 1, 2 and instrument 3 instrument. Output means 12a and input means 12b further includes an identifier, the user information / annotation input interface (i.e., input interface standard questionnaire items), integrating information format conversion module, customer / service provider platform, the expression of customer data module (i.e. , individual freedom of the input content input module), the packet temporary storage unit includes a data packaging module, encryption and decryption module, the data module decapsulating module and transmission mode selection I / O interface. Input / output devices 12a, 12b and the connection between the remote server 10 may take a combination of conventional and special corner means of communication means of communication modes of communication, as a communication mode in FIG. 7, the communication mode of communication ...... 2 8 combined mode.

In remote health Advisory System lb means 12a and 12b are designed to have a unified interface data format and design of the respective interfaces according to the different possible ft and data (inspection reports, etc.), the all possible data, image and sound formats data is organized into a unified data structure of the remote server 10 received, this modular, packaging design, to reduce the variation in the other systems of the premise, compatible to as many instruments.

b. The remote server 10

The above-described individual health analyzer 110, each group of 120 healthy and general health analysis apparatus analyzing apparatus 130 are included in the subsystem I / O modules, data management module, a data analysis module and a data storage module. In addition, the remote server further comprises means for privacy protection of personal privacy 150, management data 170 is set, the function expansion interface 180, there are alarm means, the interaction means, the mode switching ¾ health program counter and comprises a data packaging module, encryption and decryption module, the data module decapsulating module and transmission mode selection I / O interface.

(1) 150 privacy

Processing and analysis apparatus 120 of the health, the health data into the group of the Healthy apparatus 120, the health data will follow other individuals (i.e., user) put together processing, analysis and reference values ​​may be presented as to the health condition other individuals, where there is a personal privacy / groups to provide risk consulting service process is compromised by the system. Therefore, it is necessary to filter the data in some of the health information privacy, such as name, mobile phone number and so on. You can make health data in the group health before entering the first device 120 by privacy protection device, thereby filtering out about the privacy of personal / population health data in the data analysis. Health filtered data re-entry group health analysis device 120, the risk of personal / human there will not be loss of privacy.

(2) The data management apparatus 170

The remote health advice system of the first modification of the present invention further comprises means 170 analyzes an individual's health, the health data 140 in the integrity of the data management apparatus 170 to the storage device, and a remote health advisory system data management apparatus 110 transmitted , group health analysis means 120 and the data streams and general health of the health data generated by the analysis device 130 to manage the results.

The data management apparatus 170 includes a health data integrity in the memory device 140 for managing data security unit, the analysis unit, data integration unit, the data transformation unit. Wherein the data security unit is configured to collect health data security checks to stop the data into the system and carry the virus cause damage to the system.

Functional analysis unit is the health of the rendered. ^ According to sort out. Because after collecting health data 5¾ percent, the data collected is likely missing folder. For example, since forgotten, the user does not measure body weight. Such missing values ​​will affect the overall data quality, data analysis and difficult to deal with the next step. Data cleaning unit to fill the possible missing values ​​by default. For example, weight, etc., missing values, if the close time of the last measurement, the weight will be used most recently measured value as a default. And if the last measurement interval is longer, it has reached the corresponding threshold value, the judgment according to the weight measured before the trend, as a default value. Meanwhile, any default value of the IP will be marked as a reference for accurate analysis.

Data integration unit is configured to collect data on the health of the data due to various reasons the situation appear wooden compatible adjust. For example, the user may Loutian information, or use a different language with the same meaning. Data structures may be data collision, semantic heterogeneity, the need for data integration processing integrate the integrated unit.

Shen data processing apparatus 170 further comprises 1.0 to 1, the group health analysis device 120 and health data streams and data analyzer healthy individuals Laid remote health advisory system transmission means comprehensive health analysis result 130 generated manage to ensure efficient transmission of a data stream management unit data stream. In order to ensure smooth data stream, the server modularity and encapsulation, as well as system functions scalability, the system 170 (server or software) will be centrally managed, and calls the health of an individual analysis apparatus by the data management apparatus 110 generates data ( result), a group generate health data analysis means 120 and the data analysis means for generating general health, while the data management apparatus 170 in the data management unit is also responsible for flow and external data exchange (analysis report output health) management and . The data management unit 10 should be in a position in which the respective device is directly connected to a remote server system remote consultation.

The data management apparatus 170 described above into the individual functional units Falcon example only, in actual use and decomposition may be needed, or a combination of deletion, as long as the data management apparatus 170 can transmit to all remote health embodiment Advisory System data structures and data of the streaming data can be managed.

(3) Alarm device (not shown)

According to a first remote health advisory system of the second modification of the embodiment of the present invention further includes an alarm device, a method configured to detect outlier clustering algorithm based on a particular individual 140 in the previous detection and the current input of the storage means health data item health data in health; ¾ mesh data outliers, ^ a first group of a plurality of groups outlier analysis device measuring the health groups generated, and general health detection means for generating a second analysis a plurality of health data outliers groups of packets, and outlier detection of health projects ^, into line when the corresponding alarm group outliers and / or health data outliers. Alarm means outlier is detected turn to some of the data with the data or general behavior models are not consistent. This may mean that a data error, could also mean anomalies. Those skilled in the art based on what n Outlier detection method of cluster analysis, generally have the following, including those based on the statistical distribution of the outlier detection, based on the distance outlier detection, based on the density of outliers and detecting the deviation ¾ detecting outliers. Here we introduce the i ώ out this process based on outlier detection distance is used. "

In the distance-based outlier detection, if the data collection /) is at least as ^ ^ C, Γ portion of the object. Distance is greater than the threshold c / min, the object 0 is called P, C, Γ, and i / min based on the parameters of the outlier distance. That is, as long as the data set, an object does not have enough neighbors, is defined as outliers.

For example, as the data set to a two Yong age and body weight parameters, assuming a 0 data point within a radius range dmin to the number of data points is too small, the determination of which is set in advance outliers the threshold value, thus, this data point is considered an abnormal data point. While data point and the associated user and other associated data will also be identified, and notifies the user as a system record 0 ^

On the distance between the calculated data, can be based on, for example, the field of art and will be familiar Euclidean distance, Manhattan distance, Minkowski calculated from the run between each data point, the data point to the cluster, or the distance between the center between the center of the cluster, the cluster center to ^ or alarm mode referred to herein include jobs taken alarm sound, the alarm lamp or alarm pop-up window like manner. Alarm means may be located in any suitable location in the remote server 10, can be obtained as long as the above-described health data, data such as a plurality of groups on it._:.

(4) The interaction device (not shown)

Remote health advisory system 10 comprises a remote server interaction means, the interaction means that individuals provide interactive consulting exchange between parties via a remote characterized in consultation with other individuals in the system and / or health counseling services. Interactive device for all users (ie individuals), health providers and medical experts provide the appropriate exchange platforms such as BBS, chat window, website, e-mail or telephone, etc., to communicate according to different themes. For example, related service professionals and experts and users of the system and through the exchange, with a combination of system optimization and deployment, to provide better service. For example, community hospitals can do better by working with the system for the elderly community health services, individual health professionals to better understand the physical condition of its users through the system to make more timely and objective judgment. '(5) Health Project presence mode conversion means (not shown)

In remote health advisory system lb, the further comprising the presence of a feature item presence mode conversion means instance health programming mode switching means configuration mode based on the mode the instrument measurements, models and the individual standard questionnaire items free input items in frequency and a predetermined frequency comparison result of the existence of a mode-specific health programs, to determine the presence mode of the specific health item whether to switch the mode instrument measurements, the mode standard questionnaire items modes and individual freedom input items in the other one mode.

In the health data, the level in descending order according to frequency of use, including health programs exist in the mode of instrument measurements health programs, in the presence of a standard mode of questionnaire items health programs, in a pattern consisting of individual items of input-existing project. For these three use frequency level health programs, each having a frequency threshold, i.e. the predetermined frequency, when a particular order mode instrument measurement items, a model pattern standard questionnaire items modes and individual freedom input items in the present frequency health programs greater than a predetermined frequency, the conversion program to be healthy with the use of a higher frequency level, frequency level which is located in the most advanced instrument for measuring health programs in use frequency is greater than the threshold value, remains unchanged. For example, the frequency of occurrence of individual freedom than the threshold data may be input into a standard item Health Questionnaire Health Questionnaire items as a standard by being the system. When the mode of the instrument measurements, the presence of a frequency pattern model standard questionnaire items and individual free input patterns project specific medical items is smaller than a predetermined frequency, is converted into a frequency with a lower level of the health project, which is located in the lowest level of use frequency level of individual freedom in the use of health program input frequency is less than the threshold value, remains unchanged. For example, inappropriate use of frequencies below a threshold value measuring instrument health programs need not be converted to a standard questionnaire or health programs into a ho converted to individual health care input items using special instruments. Of course, when some of the health data items are not to be measuring instrument, can not be converted to the data corresponding to the health data items corresponding to the item is present in the health measuring instrument mode. Presence mode conversion data unit can set the frequency at a position of 10 is adapted to monitor the health of any data server.

Equipment

(6) Expansion Interface function 180

The remote server 10 having a function expansion interface, all future expansion unit can be connected via the interface and the system function expansion plug and play, in FIG. 7 schematically shows the function expansion unit 1, 2, 3, 4 . For example, insurance companies, insurance costs can be analyzed and designed insurance products through increased unit.

Respective components above in the second modification described during use can select any one of them or a combination of any of several different needs.

(Second Embodiment)

Hereinafter will be described in detail teleshopping diet counseling system of the second embodiment wherein the remote consultation system according to the present invention. FIGS. 10 and 11 with reference to FIG. FIG 10 is a block diagram showing a remote system according to the present shopping habits consulting Ba out a second embodiment according to Example 2, wherein "shopping habits" is another particular feature of the representation. Remote Advisory System 2 shopping habits for shopping habits condition for consultation. Including shopping habits, pursuit of the brand, with shopping preferences and so on. 10, remote shopping habits advisory system 2 includes an input device 22a, the output device 22b, the remote server 20, remote server 20 includes a storage device 240, as the analysis of individual characteristics of individual shopping habits of another example apparatus 210 analyzer , wherein the group as a group shopping habits another example of analysis apparatus analyzing apparatus 220, as an integrated feature a comprehensive analysis of the shopping habits of another example of the apparatus of the analyzer 230 and the like. The first embodiment of the present invention shown in FIG. 1 in Example 1 Comparative remote health advisory system, the system configuration of the remote shopping habits advisory system 2 with completely the same, the difference between the second embodiment and the first embodiment only wherein different data processing systems. Remote shopping habits advisory system two pairs of shopping habits characteristic data as another example of data processing, data derived relationship between the individual shopping habits as a feature project another example of a data item shopping habits and shopping habits status data, for example, the relationship between the living conditions of users (eg, alcohol consumption, exercise, occupation, etc.) and family status (such as marriage, number of children), and shopping habits (including the pursuit of the brand, with shopping preferences, etc.). FIG 11 is a schematic diagram of another example of a remote shopping habits consultation system according to the second embodiment of the present invention, two data processing display shopping habits. Here's remote shopping habits consulting either the same system can be applied to both personal shopping habits of consultation, advice can be applied to the population shopping habits. The following analysis of remote shopping habits of personal shopping habits Advisory System data will be described, remote shopping habits consulting system for analysis of the shopping habits of people only need to as a "person on the concept of" a people, shopping for the analysis of fully consistent with the data used to process in similar changes to the first embodiment, the processing of personal shopping habits of consultation with the remote system.

The following data constituting the shopping habits according to the second embodiment of the present invention will be described. 11, the same as the previously described health data, a feature of another example of data included in the shopping habits of data capable of uniquely identifying the individual items and the corresponding individual identifier an individual identifier data items, as well as a number of features shopping habits project another example of a project and the corresponding features of the project as another example of the data shopping habits project data. Shopping habits project is divided in accordance with the mode of existence, including the instruments to measure health programs, health programs and other standard questionnaire. Wherein the above-described embodiment is similar to the first embodiment, items of health measuring instrument suitable instrument is measured, for example, the supermarket checkout barcode scanner product code number identified. Standard questionnaire project is a personal shopping habits health advisory system 2 provides. In shopping habits project data, including the ability to characterize direct purchasing habits conditions include code and numeric data, such as bar code scanners from Brand, trade names such as bar code number of the reaction identified, further including the ability to have an impact on shopping habits condition data, such as occupation, family size and so on.

The following process remote personal shopping habits Advisory System 2 personal shopping habits description of the data. The input device 22a to collect data shopping habits and purchasing habits of personal data to the remote server 20.

The remote server 20 data storage device 240 according to the individual identifier habits Shopping habits ^ number of data stored thereof;

Individual shopping habits analyzing means 210 generates an individual statistical analysis and individual correlation results in accordance with the particular individual previous storage in the storage means and the current input shopping habits data, and the data based on a particular individual previous and current input shopping habits, individual system i † Analysis Studies associated with individual nodes and generate another example of the analysis results of individual characteristics as individual shopping habits analysis, 'similar to the health of an individual analysis apparatus of the first embodiment 110, in an individual shopping destroy conventional analysis apparatus 210 also includes an individual ^ gauge bridge unit, individual correlation analysis unit and the individual report generator. Statistical analysis unit performs the same processing as in the first embodiment an individual statistical analysis unit 111 and the previous particular individual data currently input shopping habits, the statistics of the current and previous particular individual shopping habits of each data item shopping habits statistical analysis of individual trends in data, resulting in changes in shopping habits reflect the trend of the situation. Self correlation analyzing unit 'performs the same analysis and correlation 112 associated with the individual analysis unit in the first embodiment Ι data,' particular individual input current and previous shopping habits find direct characterization data personal shopping habits condition shopping habits the results associated with individual association between the combination of the shopping habits of project data or the project data and shopping habits can affect the living conditions of the shopping habits of project data, such as the user's career and the pursuit of brand associations, user mood and amount of shopping association, correlation algorithm in which an individual associated with individual shopping habits of analysis equipment Tan analysis device is to focus on the implementation of the algorithm to find out the association between multiple entries set multiple transactions each firm included in the transaction consisting of multiple transactions . In association algorithms, a particular individual shopping habits and previous input data of the current transaction as a set of data for each particular individual previous shopping habits and purchasing habits of the current input data in the plurality of transaction as a transaction, and is used by at least one set of each item shopping habits Shopping habits Shopping habits of data associated with a particular data input in the previous and current data in collaboration composed of multiple items in a centralized set of items. In consideration of the case where the shopping habits of the timeliness of the data, the shopping habits of the particular individual, and previous input data of the current shopping habits including two or more data, selected from a particular individual, and previous buying habits of the current input data in pairs It Tau combined shopping habits and purchasing habits Ji cooperate to the data transaction is associated algorithm, wherein the shopping habits data in pairs selected shopping habits data input time difference to a predetermined value;: . the transaction includes all of the shopping habits of project data in pairs selected shopping habits data; a set of at least one item shopping habits and shopping habits pairs of selected data in the data composed of a plurality of items set a term set. Finally, the unit receives an individual report generation based on a particular individual, and previous shopping habits of the current input data, statistical analysis of the results of the individual and the individual associated with the results of individual shopping habits to generate analysis reports.

The following describes the group buying habits analysis device 220 pairs at a personal shopping habits data. Shopping habits analyzer 220 generates a plurality of groups (I · a plurality of groups) in accordance with all the previous individual storage ¾ opposite shopping habits and stored in the current input data ', a first plurality of statistics and a group corresponding to the group and analysis results associated with the group, and all individuals based on previous buying habits and the current input data, a first plurality of groups, the first group of the plurality of groups corresponding to the group and statistical analysis results tables generated for correlation result ^ ¾ Groups feature analysis report analyzes the shopping habits of a group of ^ a ^ instance. Embodiment of the first embodiment is similar to a remote health advisory system, remote shopping cart diet diet group Consultation System analyzer 2202 also includes a clustering unit, statistical analysis of group means, group association analysis unit and the group report generator .

Clustering unit groups buying habits analyzer 220 processes all of the individual and previous buying habits based on the current input data according to a combination of clustering method grate shopping habits of the individual data items shopping habits and purchasing habits of the items are first generated a plurality of groups, a first plurality of groups, a group of items based on a single shopping habits generated, wherein the same group is present in the shopping habits of Tau data item corresponds to a single shopping habits project shopping habits within the same data or the same data range, based on a group of a plurality of items generated shopping habits, which are present in the same group in a corresponding plurality of shopping habits of the plurality of data items ^ shopping habits thereof project data are customary identical to or within the same data range. ^ Group shopping habits gun exactly like algorithm of the first embodiment of analyzing means 220 is based clustering unit is described.

Statistical analysis Group Group shopping habits of the analysis device 220 ¥ shopping habits data element pairs each of a plurality of groups in the group statistical analysis, generating a display of a plurality of groups each group shopping habits statistical analysis of the situation of the group. Perform a correlation analysis of the following groups buying habits device 220 associated with the group analysis unit in each of a plurality of groups of the shopping habits of a group of data items shopping habits algorithm data, obtain a plurality of groups each group association shopping habits result of the combination of group data shopping habits or shopping habits project data associated with each data item shopping habits condition, the associated algorithm, using a first plurality of groups each of a group of the All data included shopping habits as a transaction set, each of a first plurality of data groups buying habits in each group of the plurality of transaction as a transaction, and using each of a first plurality of groups each data set of the shopping habits of a group of at least one of the shopping habits of project data consisting of multiple items as a set of key set. In the case of time-sensitive data to consider shopping habits,: which all individual input current and previous shopping habits included in the data belonging to the same individual shopping habits data include more than two shopping habits data, input from all previous and current individuals belong to the same individual shopping habits of the shopping habits of the data included in the data in pairs combined data selected shopping habits and shopping habits to the combined data as the association algorithm in the transaction, which, in pairs selected data input time difference shopping habits of the shopping habits of the data to a predetermined value; transaction in page I include all the shopping habits of program data selected in pairs ¾ shopping habits data; and a pair of optional shopping habits at least one set of data Ψ shopping habits project data consisting of multiple items as a set of key set.

Group and groups associated with the statistical analysis result is transmitted to the group generation unit She reported. Results group shopping habits analyzer generating unit 220 based on the first plurality of groups corresponding to the groups associated with the group and the results of the statistical analysis of the first group of the plurality of groups to generate a predetermined format was customary ^ analysis reports.

Next, 230 pairs of successive particular individual stored in the storage means 240 integrated shopping habits and analysis means clustering unit 121 when a plurality of groups buying habits Yu input data and said diet group was ¾ analyzer 220 is generated classification, prediction processing to generate a particular individual health data corresponding to the previous input multiple groups (first plurality of two groups) and the results of predictive analysis relates to the optimization of the second plurality of groups, based on the second plurality results of predictive analysis of groups and more groups involved in the second generation of a comprehensive health library optimization points ί pounds reported. Similar to the first embodiment, in the comprehensive analysis apparatus 230 shopping habits, purchasing habits comprehensive analysis apparatus similarity calculating unit 230 first calculates the shopping habits of the particular individual data Wo Π previous input current, respectively, means for generating a group of shopping habits analysis a plurality of process phases Pi similarity between a first plurality of groups each group. Comprehensive analysis device 230 shopping habits plurality of root similarity clustering section above, previous particular individual shopping habits and the current input data and processed to classify a plurality of groups selected from the plurality of specific groups a second plurality of groups of individual shopping habits and previous data corresponding to the input current. Next, to measure the page shopping habits Comprehensive analysis device 230 units Jining prediction algorithm of predicting a particular individual input current and previous shopping habits and a second plurality of data groups, and obtain a second plurality of groups multiple shopping habits correspondence group yam test results. Subsequently, a comprehensive analysis of the shopping habits of the integrated device 230 binding optimization unit similarity shopping habits of the plurality of predictive analysis results of weight assignment, to generate an optimized shopping habits results of predictive analysis. Finally, a second group of a plurality of the above-described process and to the optimization of the generated plurality of second groups into the shopping habits of prediction results report generator integrated shopping habits, purchasing habits comprehensive report generator based on a particular individual corresponding to a plurality of Groups and optimized shopping habits generate predictions comprehensive analysis of the shopping habits of another example of the analysis report as an integrated feature in a predetermined format.

The remote shopping habits advice system of the second embodiment of the present invention will be made with the same variant of the first embodiment, the data in accordance with the customary 2 shopping habits teleshopping system consulting a second modification of the embodiment It may also include individual freedom input, and further comprising a data identification means may further comprise an interaction means, the data management device, alarm device, and privacy protection unit, there is no uniform data mode switching means, an input / output device shopping habits items any one of the interface and the like, or combinations thereof.

Each processing apparatus of the present invention described above were used in various algorithms including clustering, correlation algorithm prediction and similarity calculation algorithm, a clustering algorithm for detecting outliers is based skilled etc. those well-known in relation to its specific principles and algorithms process can be found in Data Mining Concepts and Techniques, Second Edition by [Canada] Jiawei Han, Micheline Kamber.

The various embodiments described above the remote health advisory system in accordance with an example of a feature of the remote consultation system according to the invention and a remote shopping habits advisory system, in addition to the health of individuals or populations. Other features of the situation than the situation of the shopping habits of consultation system can also be used as an example feature remote consultation system. In addition, for simplicity of illustration, according to the instructions remote health advisory system embodiment of the present invention and remote shopping habits of consultation system, only given a detailed explanation to solve the key components of technical problems to which the present invention is required, and explanation is omitted for other non-critical components such as the display system, the power supply device and the like.

Additional advantages and modifications to the skilled person are conceivable. Accordingly, the present invention in terms of its broader aspects, is not limited to the specific details shown and described with exemplary embodiments and examples. In the case where the overall spirit and scope of the invention without departing from the spirit of the appended claims and their equivalents define the concept, various modifications may be made.

Claims

Rights request
An advisory system wherein remote characteristic condition for consulting, characterized by comprising:
Input means for collecting data input features, said features comprising a data item identifier and the individual identifiers of individual data items corresponding to the item number and the characteristics of the characteristics of data items corresponding to the item characteristic; and
Remote server, including
Storage means storing the data according to the characteristic data of the individual identifier feature data items;
Wherein the individual analysis apparatus, and generates the individual statistical analysis according to a specific individual associated with the individual results of said storage means stores the previous and the current input feature data, and based on the previous and the particular individual characteristic data currently input, the individual statistical the results and the individual characteristics of individual correlation result generating analysis report, the group feature analysis means, generates a first plurality of groups based on all the previous individual stored in said storage means and the current input feature data, and the first group and groups associated with the statistical analysis results corresponding to the plurality of groups, and statistical analysis results and the group associated with the group generated based on analysis of the first feature plurality of groups, the group, and
Comprehensive analysis apparatus wherein said generating means generating a first group of the plurality of previous inputs to the particular individual based on said analysis stored in said storage means a particular individual features of previous and current input data and wherein the group feature data corresponding to a second plurality of said second group and to a plurality of optimization results of predictive analysis group, and generates a plurality of integrated features based on the second group and the optimized prediction analysis result report.
Wherein said remote consultation system as claimed in claim 1, characterized in that, further comprising:
Output means configured to receive and analyze the output of the report in a predetermined format individual characteristics, and the group wherein the composite analysis feature analysis.
3. The remote characterized in consultation system according to claim 1, wherein said analyzing means comprises individual characteristics: Individual ^ count analysis unit, and configured to store the characteristics of the specific subject means the previous input current statistical analysis of the individual data for statistical analysis, generating trends reflect the characteristics of the condition;.
Self correlation analyzing unit configured to obtain a single algorithm wherein said data items a particular individual characteristics of the current and previous input data, the characteristic data of a particular individual current and previous input feature data item associated with the implementation of feature data items the individual correlation result associated data item permutations and features are characteristic of the condition; individual and report generation unit configured to, based on the previous and the particular individual characteristics of the current input data, and the individual results of the statistical analysis said correlation result to generate the individual features of the individual analysis.
4. The remote characterized advisory system according to claim 3, characterized in that the individual characteristic analysis performed by the correlation algorithm associated with individual device to identify the analysis unit is concentrated in a transaction composed of a plurality of the transactions algorithm for the association between the plurality of item sets in each of a plurality of transaction transaction contains.
5. The remote characterized advisory system according to claim 4, wherein, in said correlation algorithm, using the specific subject characteristic data and previous input current as a transaction set, the previous and current input the particular individual each feature data in the feature data of said plurality of transaction as a transaction, using a set of at least one characteristic feature of each data item of said particular individual characteristic data of the current and previous input data composed of said plurality of items as a set of items set.
6. The remote characterized advisory system according to claim 4, characterized in that dichotoma, wherein the feature data further comprises a data input time, to the particular individual characteristic data and previous input current comprises two or more said characteristic data, from the previous particular individual characteristic data and the current input in pairs characteristic data combinations, and in that said feature data as the transaction associated algorithm, wherein, in pairs selected feature wherein the data input time difference to a predetermined value;
The transaction data includes all Patent ΪΕ items in pairs said selected feature data; and wherein the at least one data item of the selected pair of feature data as a set composed of a plurality of items concentrated a set of items.
7. The remote characterized in consultation system according to claim 1, wherein said group comprising a dispensing feature - a clustering unit configured to respectively based on the individual characteristics of the storage means and a plurality of feature item item performing clustering the feature data of the previous and current input all individuals to generate a first plurality of the group, in said first plurality of groups,.:.
Corresponding to a group based on a single feature item generated, wherein the presence in the same group or the same feature data within the same range of data items to the single characteristic feature item data,
Characterized by a plurality of group items based on the generated, wherein the same group is present in the feature data items corresponding to the plurality of features noted in the number of the plurality of features are the same items or data, respectively, within the same range; statistical analysis of the group unit, configured to the first feature data of each of a plurality of groups of the statistical analysis group, characterized in generating a display condition of the first plurality of groups each group statistical analysis of the group
Group association analysis unit, configured to perform a bone-related feature item data in the first data characteristic of each of the plurality of groups of a group of algorithms, each of said derived first plurality of groups the result of the combination group association group characteristic feature data or feature data items are data items associated with the feature condition; and
Group report generation unit configured to, based on the first plurality of groups, wherein the groups of the statistical analysis, and the result generating analysis group association.
8. The remote characterized advisory system according to claim 7, wherein said analyzing said feature group associated with the group association means algorithm executed in the analysis unit is focused to identify multiple transactions in a transaction consisting of algorithm for the association between the plurality of set items of each transaction containing said plurality firm.
9. The process of claim 8 wherein advisory system as claimed in claim, characterized in hand, the operator in association method, using all the features of a plurality of data groups each of a group included as a the transaction set, each characteristic data of each of the first group of the plurality of groups of said plurality of transaction as a transaction, using one of said first plurality of groups each of at least one characteristic of each item of data in the feature data group as a set consisting of a plurality of items of the itemsets set.
Ij to 10 weight Wo - seeking remote consultation system of claim 8 wherein, wherein the characteristic data include characteristic data into a ho input time, the previous and current input all individuals ^ features included in the data belongs to the same individual feature data comprising two or more said characteristic data,
Belonging to the same characteristic data from said individual comprising the individual all previous and current input feature data in the feature selected in pairs for data combinations, and in that characteristic data as the transaction associated algorithm, wherein feature data in pairs selected feature data input time difference to a predetermined value;
The transaction data includes all of the features of the items in pairs selected Laid ^ data; and wherein the at least one data item of the selected pair of feature data as a set composed of a plurality of items of the fan concentrated a set of items.
1 1. The remote characterized in consultation system according to claim 1, wherein said integrated feature analyzing apparatus comprising: a similarity calculation unit configured to calculate said storage means to reflect the particular individual stored in the previous and wherein the current input data are analyzed with the group wherein a plurality of similarity degree of similarity between the first means for generating a plurality of groups each group a;,
Collation unit configured to, based on the plurality of similarity of the characteristic data of the particular individual and previous input current and said first plurality of groups are classified from the first process to a plurality of groups a group selected from the group of said second plurality of previous and particular individual characteristic data corresponding to the input current;:
Prediction algorithm unit configured to subject the specific features of previous and current input data and a plurality of said second group of prediction algorithms to derive predicted execution condition characterized in correspondence with the second group of the plurality of change a plurality of results of predictive analysis;
Integrated optimization unit configured to perform according to the similarity of the plurality of the plurality of weight distribution results of predictive analysis results of predictive analysis to generate said optimized; and
Comprehensive report generation unit configured analysis report based on the second plurality of groups and said climbing optimization generate the composite results of predictive analysis feature.
12. The remote characterized in consultation system according to claim 1, characterized in that, further comprising an alarm means,
'Configured based clustering algorithm ^ outlier detection method for detecting the feature item in said storage means a particular individual characteristic data of the current and previous input data of the number of items characterized Chi outlier,' Detection analyzing said outlier group wherein the group is set to generate a first plurality of ¾ of the group ^, wherein said integrated analyzing said detecting means for generating a second plurality of feature data included in the group outlier point, and the feature item data found outlier, the corresponding alarm group and / or when the feature data outliers outlier.
As claimed in any of claims 1-12 wherein the remote consultation system according to claim, characterized in that said characteristic pattern data included in each item instrument for measuring the presence of items and features present in the standard mode, characterized in questionnaire items project. .
14. The remote characterized advisory system according to claim 13, wherein,
The remote server further comprises data identifying means, wherein said data when said input means includes collecting individual freedom of input, the input content data identification means consisting of additional features identified from the individual items and the corresponding feature item data, and additional features added to the program and data of the corresponding characteristic feature data items, said additional feature f household items present in a pattern consisting of individual items in the input characteristic number ¾ in.
In the meter mode based on measurements of ^, said standard pattern and the feature item questionnaire items presence mode conversion means arranged: 15. The remote advisory system characterized according to claim 14, characterized in that, further comprising using a predetermined frequency of the frequency comparison result of the presence of a pattern of said pattern consisting of individual items of input items in a particular feature, there is a particular feature of the mode determination whether the item is converted into the mode of instrument measurements, the standard questionnaire Further projects a pattern model and the pattern consisting of individual items of input.
16. The remote characterized advisory system according to claim 14, characterized in that, further comprising:
Interaction device, such that the interaction means provides the individual with other individuals and / or advisory services interactive consulting wherein communication between remote parties via said system characterized in consultation.
17. The remote characterized advisory system according to claim 14, characterized in that, further comprising - a data management apparatus, wherein the data integrity of said data management means manages said storage set in, and wherein the remote system consulting the individual features of the transmission analyzing means, said analyzing means and said group of features comprehensive characterization data streams generated by means of the results and the feature data.
18. The remote characterized advisory system according to claim 14, characterized in that, further comprising:
Privacy protection device, a privacy device enters the group characterized in that said filter characteristic data privacy information of the characteristic data prior to said analyzing means to prevent leakage of private information.
19. The remote characterized advisory system according to claim 14, wherein said output means and said input means having a unified interface data format.
20. A method for remote consultation characteristic feature Advisory condition, characterized by comprising:
The input processing, the input feature data collected, the feature data item identifier and including individual data items corresponding to the individual identifier, a number of characteristic data items and the features and feature item corresponding to the item;
Storage processing according to the characteristic data of the individual identifier wherein the data storage data items; individual characteristics analysis process to generate the statistical analysis according to a particular individual subject of previous processing in the storage and the feature data stored in the current input and the individual correlation results, and based on the particular individual calendar ^ and the current input feature data, and the individual results of the statistical analysis of the correlation result to generate the individual characteristics of individual analysis, wherein the evaluation group, the storage process in accordance with All individual characteristic data and stored previous input current to generate a first plurality of groups, group results of statistical analysis and the results of the group associated with the first plurality of groups corresponding to the plurality of groups based on the first and group, statistical analysis of the group associated with the group and the group result generating analysis features, and
Comprehensive characteristics analysis process, the analysis process generates a plurality of a first group to generate the particular individual based on said input storage previous process stored in a particular individual characteristic data and previous input current and wherein the group feature data corresponding to a second plurality of said second group and to a plurality of optimization results of predictive analysis group, and generates a plurality of integrated features based on the second group and the optimized prediction analysis result report.
21. The method of consulting a remote characterized according to claim 20, characterized in that, further comprising - an output processing, the receiving and analysis report in a predetermined format to output the individual features, and the group wherein said integrated feature analysis Analysis report.
22. The method of consulting a remote characterized in claim 20, wherein the individual feature analysis process comprising: a step of statistical analysis of an individual, wherein said data stored in said storing process specific subject of the current input and previous the statistical analysis, generating trends reflect the characteristics of the individual situation of the statistical analysis;
Self correlation analyzing step of performing a data item associated with the particular individual features of previous and current input data in the derived algorithm wherein a single feature data item and the particular individual characteristics of the current input the previous data, the program data and combination of features the individual correlation result associated data items are arranged in characteristic features of the condition; and
Individuals report generation step, based on the particular individual and previous characteristic data currently input, the subject and the individual results of the statistical analysis of the correlation result to generate individual feature analysis.
23. The method of consulting a remote characterized in claim 22, wherein said individual wherein said analyzing algorithm associated individual processing steps associated with the analysis performed to identify the transaction is to focus a transaction composed of a plurality of said algorithm for the association between the plurality of item sets in each of a plurality of transaction transaction contains.
24. The method of consulting a remote characterized in claim 23, characterized in that the correlation algorithm, using the specific subject characteristic data and previous input current as a transaction set, the previous and current input the particular individual each feature data in the feature data of said plurality of transaction as a transaction, using a set of at least one characteristic feature of each data item of said particular individual characteristic data of the current and previous input data composed of said plurality of items as a set of items set.
25. The method of consulting a remote characterized in claim 23, wherein the feature data further includes feature data input time, to the particular individual characteristic data and previous input current comprises two or more said characteristic data,
From the previous particular individual characteristic data and the current input in pairs characteristic data combinations, and in that characteristic data as the algorithm associated with the transaction in which, in pairs selected wherein the data input time difference of the feature data into a predetermined value, the transaction data includes all of the features of the items in pairs in a selected feature data; and
Wherein the at least one data item in pairs in the selected feature data as a set composed of a plurality of items in a sink itemsets.
26. The method of consulting a remote characterized in claim 20, wherein said group characteristic analysis process comprising: a clustering step, respectively, and a plurality of items based on a single feature of the feature item storage processing for all the individual features of the current input data and previous execution of the clustering algorithm to generate a first plurality of groups, said first plurality of groups,
Corresponding to a group based on a single feature item generated, wherein the presence in the same group or the same feature data within the same range of data items to the single characteristic feature item data,
Wherein the plurality of data items from the plurality of groups based on the generated feature item, wherein, in the same group is present in the data corresponding to the plurality of characteristic features are the same as or items are within range of the same data;
Group statistical analysis step, the characteristic data of said first plurality of groups each group for statistical analysis, generate a feature for each state of displaying the first group of the plurality of groups said statistical analysis group; group association analysis step, performing a first correlation of the feature item of each of a plurality of groups of features in the group data algorithm, deriving the first plurality of groups each of the combined groups characteristic feature data or feature data item of the data items associated with the group association results are characteristic of the condition; 'and
Group report generation step, based on the first plurality of groups, the group the group and statistical analysis to generate the correlation result fans Laid Zheng group analysis.
27. The method of consulting a remote characterized in claim 26, wherein said analyzing said feature group associated with group association algorithm processing analysis step is performed in a transaction composed of a plurality of transactions to find concentrated algorithm associated to the set of transaction each transaction contains a plurality of said plurality of items.
28. The method of consulting a remote characterized in claim 27, characterized in that all the associated characteristic data using the first algorithm ^ plurality of groups each group included as one of the transaction set, each characteristic data of each of the first group of the plurality of groups of said plurality of transaction as a transaction, and using each of a first group consisting of said plurality of groups of the at least one characteristic feature of each data item in the data set consisting of a plurality of items as a set of items set.
29. The method of consulting a remote characterized in claim 27, wherein the characteristic data include characteristic data into a ho input time, all the previous individual characteristic data and the current input feature data included in the same individual belonging to comprising two or more said characteristic data,
Belonging to the same characteristic data from said individual comprising the individual all previous and current input feature data in the feature selected in pairs for data combinations, and in that characteristic data as the transaction associated algorithm, wherein , wherein the difference in pairs of the selected feature data input time data of a predetermined value, the transaction data includes all of the features of the items in pairs in a selected feature data; and
In pairs selected Laid ί! Characterized at least one data item as a set of data consisting of & ¾ of the item set a plurality of set items.
30. The method of consulting a remote characterized in claim 20, wherein said integrated feature analysis process comprising: ho similarity calculating step, calculating the storage process reflecting the particular individual stored in the previous and current input a plurality of respective feature data of the degree of similarity between the similar process to generate a first plurality of groups each group with said group characteristic analysis;
Grouping step, based on the plurality of similarity of the characteristic data of the particular individual and previous input current and said first plurality of groups classified to select from a plurality of said first group the particular subject that the previous and the current input feature data corresponding to a second plurality of groups;
Ho step prediction algorithm, the previous Patent ¾ individual characteristic data and the current input and the second group of the plurality of prediction algorithms perform status changes to derive prediction feature and the second group of the plurality of one- a plurality of results of predictive analysis; comprehensive optimization step of performing a plurality of weights based on the similarity of said plurality of results of predictive analysis reallocation results of predictive analysis to generate said optimized; and
Synthesis report generation step, wherein said integrated analysis report is generated based on the second plurality of groups and the optimization of results of predictive analysis.
31. The method of consulting a remote characterized in claim 20, characterized in that, further comprising an alarm processing method for clustering algorithm based on the detection of outliers, the specific subject detecting previous and current input feature data in the feature feature item data outlier data item, detecting the outlier group wherein said group analysis process of generating a first plurality of groups, detecting the comprehensive characterization of the process of generating a second characteristic data of a plurality of outlier included in the group, and found that the feature item data outliers, the group corresponding alarm and / or when the feature data outliers outlier.
32. any of claims 20-31 wherein the remote consultation method according to, characterized in that said characteristic pattern data included in each item instrument for measuring the presence of items and features present in the standard mode, characterized in questionnaire items project.
Wherein the remote consultation 33. The method according to claim 32, wherein,
Ho into the remote server comprising a data identification processing, when the input processing said feature data collected includes individual freedom input, processing the data to identify the individual free from the additional input of the content identification and the corresponding feature item the feature item data, and additional features added to the program and user data of the corresponding feature item; F said characteristic data, the additional features in the feature data item subject to the free page for the input pattern I is present.
34. The method of consulting a remote characterized in claim 33, characterized in that, further comprising:
Wherein the presence mode conversion processing program, based on a comparison of the frequency of use in a mode of instrument measurements, a standard questionnaire Mode Mode Mode the individual item and the free input items in presence of a particular feature item with a predetermined frequency, as a result, determining the presence of specific features of the pattern project to see if the instrument is converted to mode measurement items, a further mode of the mode pattern standard questionnaire item and the individual freedom of input items.
PCT/CN2010/000431 2009-04-03 2010-04-02 System and method for remote feature consulting WO2010111895A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN 200910048809 CN101853428A (en) 2009-04-03 2009-04-03 Remote characteristic query system and method
CN200910048809.0 2009-04-03

Publications (1)

Publication Number Publication Date
WO2010111895A1 true WO2010111895A1 (en) 2010-10-07

Family

ID=42804901

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/000431 WO2010111895A1 (en) 2009-04-03 2010-04-02 System and method for remote feature consulting

Country Status (2)

Country Link
CN (1) CN101853428A (en)
WO (1) WO2010111895A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019113323A1 (en) * 2017-12-06 2019-06-13 Cardiac Pacemakers, Inc. System to improve clinical workflow

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104932966B (en) * 2015-06-19 2017-09-15 广东欧珀移动通信有限公司 Detecting an amount of brush application downloading method and apparatus
TWI578262B (en) * 2015-08-07 2017-04-11 Wistron Corp Risk assessment system and data processing method
CN106580282A (en) * 2016-10-25 2017-04-26 上海斐讯数据通信技术有限公司 Human body health monitoring device, system and method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060036619A1 (en) * 2004-08-09 2006-02-16 Oren Fuerst Method for accessing and analyzing medically related information from multiple sources collected into one or more databases for deriving illness probability and/or for generating alerts for the detection of emergency events relating to disease management including HIV and SARS, and for syndromic surveillance of infectious disease and for predicting risk of adverse events to one or more drugs
US20070248986A1 (en) * 2006-04-10 2007-10-25 Academisch Ziekenhuis H.O.D.N. Lumc Systems and methods for predicting an individual's risk of developing rheumatoid arthritis

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060036619A1 (en) * 2004-08-09 2006-02-16 Oren Fuerst Method for accessing and analyzing medically related information from multiple sources collected into one or more databases for deriving illness probability and/or for generating alerts for the detection of emergency events relating to disease management including HIV and SARS, and for syndromic surveillance of infectious disease and for predicting risk of adverse events to one or more drugs
US20070248986A1 (en) * 2006-04-10 2007-10-25 Academisch Ziekenhuis H.O.D.N. Lumc Systems and methods for predicting an individual's risk of developing rheumatoid arthritis

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019113323A1 (en) * 2017-12-06 2019-06-13 Cardiac Pacemakers, Inc. System to improve clinical workflow

Also Published As

Publication number Publication date
CN101853428A (en) 2010-10-06

Similar Documents

Publication Publication Date Title
Morillo et al. Interdisciplinarity in science: A tentative typology of disciplines and research areas
Huysmans et al. An empirical evaluation of the comprehensibility of decision table, tree and rule based predictive models
Freifeld et al. HealthMap: global infectious disease monitoring through automated classification and visualization of Internet media reports
Hardisty et al. A dirty word or a dirty world? Attribute framing, political affiliation, and query theory
Luo et al. Building association link network for semantic link on web resources
Atzmueller Subgroup discovery
Fawcett An introduction to ROC analysis
Friemel The digital divide has grown old: Determinants of a digital divide among seniors
Markines et al. Evaluating similarity measures for emergent semantics of social tagging
Yu et al. The impact of social and conventional media on firm equity value: A sentiment analysis approach
Cosco et al. Lay perspectives of successful ageing: a systematic review and meta-ethnography
Coussement et al. Improving customer complaint management by automatic email classification using linguistic style features as predictors
Hook et al. New fathers? Residential fathers’ time with children in four countries
Matsuo et al. POLYPHONET: an advanced social network extraction system from the web
Borgman et al. Scholarly communication and bibliometrics
Mingers et al. A review of theory and practice in scientometrics
Li et al. Using text mining and sentiment analysis for online forums hotspot detection and forecast
Rubin et al. Veracity roadmap: Is big data objective, truthful and credible?
Sugimoto et al. Scholarly use of social media and altmetrics: A review of the literature
Baggio et al. Quantitative methods in tourism: A handbook
Wildgaard et al. A review of the characteristics of 108 author-level bibliometric indicators
Gandomi et al. Beyond the hype: Big data concepts, methods, and analytics
Brew et al. Using Crowdsourcing and Active Learning to Track Sentiment in Online Media.
David et al. Finding frames: Comparing two methods of frame analysis
Chakraborty et al. Text mining and analysis: practical methods, examples, and case studies using SAS

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 10757998

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase in:

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 17/02/2012)

122 Ep: pct application non-entry in european phase

Ref document number: 10757998

Country of ref document: EP

Kind code of ref document: A1