CN1973281A - Method and apparatus for survey processing - Google Patents
Method and apparatus for survey processing Download PDFInfo
- Publication number
- CN1973281A CN1973281A CN 200480013206 CN200480013206A CN1973281A CN 1973281 A CN1973281 A CN 1973281A CN 200480013206 CN200480013206 CN 200480013206 CN 200480013206 A CN200480013206 A CN 200480013206A CN 1973281 A CN1973281 A CN 1973281A
- Authority
- CN
- China
- Prior art keywords
- data
- investigation
- enquiry
- survey
- enquiry data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Certain embodiments of the present invention provide a system and method for improved, real-time survey data processing and analysis. Certain embodiments use distributed computing techniques to perform statistical calculations for survey data, such as general purpose and multi-rater feedback data. Certain embodiments use a matrix structure coupled with hash tables for efficient processing of survey statistics across an entire survey. Certain embodiments use a criteria parser based on externally definable lexical rules to determine which survey responses belong to which groups. Certain embodiments allow coalescing of summarization requests between summarization servers to allow for peak performance across any number of surveys. Certain embodiments provide a scalable, adaptable survey processing system and method that processes survey results in real time, allowing for immediate feedback during the survey process.
Description
The present invention relates to the U.S. Provisional Application No.60/471 of " method and apparatus that is used to investigate " by name of submission on May 16th, 2003,223, and require its right of priority.
Technical field
The present invention relates generally to that enquiry data is handled and analysis.More specifically, the present invention relates to improved real-time enquiry data handles and analyzes.
Background technology
A lot of companies pay attention to the investigation of for example employee's assessment and satisfaction investigation.For example, company uses the employee to assess to judge performance and promotes.Company also uses customer satisfaction survey to weigh the success of products ﹠ services, and determines to improve.Use investigation, manager can also be assessed by the employee.
For example, investigation can help to start the variation in work place environment, product or service improvement and the employee's training.Investigation result may have influence on the great strategic decision of company.Current investigating system can not provide fast response to help make important, commercial decision-making timely.And current investigating system is difficult to visit as required.
Therefore, providing the real-time management of investigation and the system of processing will be unusual desirable.And, improved system to the accessibility of investigation, enquiry data and investigation statistics and will be very desirable.The system that investigates that allows investigation information and classification to be dynamically adjusted also will be unusual desirable.Therefore, exist being used for the demand of the system and method that improved real-time enquiry data handles and analyze.
Summary of the invention
Some embodiments of the present invention provide and have been used for the system and method that improved real-time enquiry data is handled and analyzed.Some embodiment use distributed computing technology to carry out for example general purpose and the statistical computation of the enquiry data of feedback data (multi rater feedback data) at many levels.Some embodiment use the matrix structure with the Hash table coupling, are used for effectively handling the investigation statistics that runs through whole investigation.Some embodiment use based on the standard analysis device (parser) of outside definable semantic rules and determine which group which investigation response belongs to.Some embodiment allow multiplexing web feed request between each summary server, run through the peak performance of the investigation of any amount with realization.Some embodiment provide scalable (scalable), the adaptive system and handle the method for investigation result in real time of investigating, and allow the immediate feedback during fact-finding process.
In one embodiment, the system of investigating comprises: the read engine adapter is used to read enquiry data and the mark starting point to the processing of enquiry data; The survey processor is used to use semantic analysis to handle enquiry data, and described survey processor produces statistics according to enquiry data at wherein at least one of investigation, problem and scope; With, write the engine adapter, be used to export the investigation statistics of renewal.
Read engine can be read previous wrap-up state from database, so that be that original state is set up in summary and statistical treatment.In one embodiment, the survey processor comprises the standard analysis device, is used for according to the definable semantic rules enquiry data being sorted out.The survey processor can identify the report group that enquiry data is its member.System can also comprise the Hash matrix, is used to store the statistics to investigation.
In one embodiment, investigation and analysis system in real time comprises: run time engine is used to receive enquiry data; At least one summary also closes device, is used for the Classification Count data and enquiry data is routed at least one statistical abstract engine; With, at least one statistical abstract engine, be used for analysing and investigating in real time data with executive summary and statistical study wherein at least one.Described at least one summary also closes device and classifies and the route enquiry data according at least one standard.
In one embodiment, described at least one summary and close device and receive enquiry datas stream from a plurality of sources.Enquiry data can comprise investigating finishes data and/or new survey report group data.One or more standard can comprise that one or more allows summary and closes the standard that device divides into groups according to the optimal ordering of handling to the enquiry data that enters (incoming).In one embodiment, making a summary and closing device compares the history list of a plurality of enquiry datas with recently processed investigation, so that described a plurality of enquiry datas are routed at least one statistical abstract engine.Described at least one or more a plurality of statistical abstract engine can read previous wrap-up state from database, so that calculate the mark starting point for the accrual accounting of using enquiry data.
In one embodiment, implementing the investigation and analysis system also comprises: report engine is used to produce new group that is used for survey report.System can also comprise database, be used to store enquiry data and survey report group the definition wherein at least one.In addition, system also comprises user interface, is used to import enquiry data.In one embodiment, user interface is based on the user interface of network (Web-based).
The embodiment of method that is used for the real-time processing of enquiry data comprises: receive and search into a matter; Determine the described route that searches into a matter according at least one standard; Described searching into a matter is routed to the device of investigating; Device is handled described searching into a matter in real time investigating; With, the real-time feedback that searches into a matter from described is provided.This method can also comprise determines whether the incident of entering comprises enquiry data or new investigation group.In one embodiment, treatment step comprises the statistical treatment to enquiry data.Treatment step can also comprise creates new investigation group.In one embodiment, treatment step is analysed and investigated response according to investigation, problem and/or theme.This method can also comprise according to processed searching into a matter upgrades the data of being stored.
Description of drawings
Fig. 1 shows the investigating system that uses according to embodiments of the invention.
Fig. 2 shows the investigation and analysis system that uses according to embodiments of the invention.
Fig. 3 shows the statistical abstract engine that uses according to embodiments of the invention.
Fig. 4 shows and is used at the process flow diagram of the statistical abstract engine that uses according to embodiments of the invention to the method for event message classification.
Fig. 5 reads the process flow diagram of the method for web feed request when showing and preparing for the summary that uses according to embodiments of the invention.
Fig. 6 shows the process flow diagram of the method that is used for the statistical abstract used according to embodiments of the invention.
Fig. 7 shows according to embodiments of the invention, is used to produce the report layout system that shows the investigation and analysis data document.
Fig. 8 shows the process flow diagram that is used for e-survey management of using according to embodiments of the invention and the method for analyzing.
When read in conjunction with the accompanying drawings, the summary of the invention and the following detailed description to some embodiments of the present invention of front will be understood better.For purpose of the present invention is described, show some embodiment in the drawings.But, should be appreciated that configuration and the means shown in the present invention is not limited in the accompanying drawings.
Embodiment
Fig. 1 shows the investigating system 100 that uses according to embodiments of the invention.System 100 comprises user interface 110, database 120, investigation and analysis system 130 and exports 140.
For example, in one embodiment, user interface 110 can be based on the interface of network or other data input form.Can be with for example HTML(Hypertext Markup Language), extend markup language (XML), Simple Object Access Protocol (SOAP), perl, java or other CGI (Common Gateway Interface) (CGI) script are implemented user interface 110, to catch the user investigation response, as the data of database 120.Use for example transmission of messages of Microsoft's message queue, data and other message can be delivered to database 120 from user interface 110.In one embodiment, database 120 is operating structure query language (Structured Query Language, SQL) servers, and be configured to the server that data provide the four-processor server cluster of redundancy and security for example.User interface 110 can reside on the computing machine identical with database 120 or investigation and analysis system 130, perhaps may be implemented within on the independent computing machine.For example, user interface 110 can be moved on uniprocessor or Dual Processor Server.The parts of system 100 can be networked by connection, wireless connections or other wired connection by for example Ethernet, gigabit Ethernet.In one embodiment, use the procotol of TCP/IP as communication between the parts of system 100.For example in one embodiment, investigation and analysis system 130 is four-processor servers.The parts of system 100 may be implemented within software and/or the hardware.
Fig. 2 shows the investigation and analysis system 200 that uses according to embodiments of the invention.Investigation and analysis system 200 and top similar with reference to the described investigation and analysis of Fig. 1 system 130.Analytic system 200 comprises investigation run time engine 210, survey report engine 220, makes a summary and closes device 230-233 and statistical abstract engine 240-243.
Investigation run time engine 210 receives enquiry data from user interface 110.Enquiry data is sent to database 120.Investigation run time engine 210 is also to making a summary and closing one of them transmission investigation of device 230-233 and finish message.Similarly, survey report engine 220 also defines to database 120 transmission groups, and to making a summary and closing one of them transmission group definition of device 230-233 and finish message.Summary also closes device 230-233 and statistical abstract engine 240-243 communication, so that sum up summary and carry out enquiry data for example or the statistical study of group definition.Statistical abstract engine 240-243 is from database 120 reception information and new database 120 more.
In one embodiment, after user interface 110 places have finished for example investigation of based on network investigation, investigation run time engine 210 produces investigation and finishes incident the user.This incident comprises the information of sign investigation response record, and for example, the investigation response record is the form of database identifier, investigation identifier and user answer person's identifier combination.Investigation is finished incident and can be sent straight to statistical abstract engine 240-243, perhaps is sent to summary and closes device 230-233 to be used to be routed to suitable statistical abstract engine 240-243.The incident of finishing of making a summary can also be marked as the request that is not done to user's summary in database 120.
Similarly, in another embodiment, the user can create new group for report.Survey report engine 220 produces new group of incident, newly organizes the identifier of this group of event report.New group incident can be sent straight to statistical abstract engine 240-243, perhaps is sent to summary and closes device 230-233, is used to be routed to suitable statistical abstract engine 240-243.Survey report engine 220 can also send new group definition to database 120, is used for using when building new summary record for example forming for this.
Summary also closes device 230-233 from a plurality of sources reception flows of event.Flow of event comprises for example newly organizes definition incident and/or the new incident of finishing of organizing.Summary also closes device 230-233 the message that enters is divided into groups, and for example by optimal ordering message is divided into groups, to help effectively to sum up summary.Message is sent to statistical abstract engine 240-243 by the group of response (finishing incident for investigation) being made in which investigation according to the respondent or just be defined (for new group of definition incident).
Fig. 3 shows the statistical abstract engine 3 00 that uses according to embodiments of the invention.Statistical abstract engine 3 00 and top similar with reference to the described statistical abstract engine of Fig. 2 240-243.Statistical abstract engine 3 00 comprises engine adapter (reading) 310, statistical treatment device 320 and engine adapter (writing) 330.The parts of digest engine 300 are described below in more detail.
Summary also closes device 230 reception new events.Summary also closes the message that enters in device 230 classification queues.Summary also closes device 230 and also according to optimal ordering the message that enters is divided into groups to help effective summary.In one embodiment, making a summary and closing device 230 has optimization aim, comprises according to the current queue state, by possible maximum group summary information is divided into groups, and guarantees that each investigation is given " concern ".For example, Duo 1000 times investigation and submit incident to, investigate also that unnecessary wait is the long time of B with analyzed and renewal even investigation A has than investigation B.In one embodiment, make a summary and close device 230 and move always, till in message queue, not having message.When formation when being empty, make a summary and close device 230 " dormancy ", till receiving new information.If formation is never empty, then makes a summary and close device 230 and can move constantly.
Fig. 4 shows and is used at the summary that uses according to embodiments of the invention and closes device 230 to event message classification or and the process flow diagram 400 of the method for closing.In one embodiment, make a summary and close the incident that device 230 processing as much as possible belong to investigation.At first, in step 410, at summary and close device 230 and receive new message notification from message transmission (messaging) system of for example Microsoft Message Queue (Microsoft's message queue).Then, in step 420, make a summary and close device 230 and determine investigation to be processed.Then, in step 430, make a summary and close device 230 and check record in the formations restriction of " t " individual element (for example, up to).In step 440, remove the record that belongs to the investigation that is selected for processing from formation.Then, in step 450, be grouped from the data of incident, and be passed to digest engine 300.
As mentioned above, make a summary and close device 230 and determine to handle which investigation.In one embodiment, make a summary and close the history list that device 230 remains with recently processed investigation.In order to determine to handle which investigation, make a summary and close device 230 by finding the candidate list of editing possible combination in all unique investigation of in the formation of certain degree of depth " d ", being represented.In case formed candidate list " C ", then candidate list is compared with historical classification.History list " H " representative processed investigation recently is up to the degree of depth " h ".Record in the history list can be listed by the order of finishing, and the investigation that for example is done recently is at the top.
In one embodiment, make a summary and close device 230 and attempt in " C ", to find first not record in " H " (C/H), and calculate ordered set " HC " simultaneously, the common factor (H ∩ C) of set " HC " representative " H " and " C ".If (C/H) evaluation obtains nonempty set, then first from (C/H) is selected as summary candidate person.If (C/H) evaluation obtains empty set, then from selecting candidate by the last element the set (H ∩ C) of " HC " representative.If this set is a non-NULL, then first candidate is not also had processed candidate list to take out from recent.If this set is empty, then all candidates are processed recently.Like this, represent the candidate (from last element of (H ∩ C)) of recently processed investigation processed.In case determined messaging list, then this tabulation is edited as set " M ", and set " M " can be represented by the vector structure in inside.Then, this vector is passed to digest engine 300 and is used for handling.
As mentioned above, the quantity of " t " representative clauses and subclauses of " searching (peek) " formation when calculating " M " massage set of in the single iteration of statistical abstract engine 3 00, wanting processed.In one embodiment, " t " is provided with according to available RAM on the computing machine of operation statistical abstract engine 3 00.For example, the server with 2GBRAM can be handled from about 100,000 investigation of single investigation and finish incident.
The quantity of candidate search depth " d " representative clauses and subclauses of " searching " formation when calculating the candidate investigation set " C " of wanting processed.The degree of depth " d " helps to guarantee at summary and closes investigating in the example of device 230 when mainly being a spot of investigation, makes a summary and closes device 230 and do not cost a lot of money the time on the collection candidate.
The maximum quantity of the candidate incident that set of candidates restriction " c " indication will be collected.For example, reach the degree of depth " d " in the formation and produced " c " individual clauses and subclauses before if candidate is collected to operate in, then candidate is collected operation cycle and is stopped.When a large amount of different investigation had been handled simultaneously, restriction " c " helped to optimize the time that is spent on the candidate collecting.
The maximum quantity of the candidate history entries that historical set restriction " h " indication will be remembered." h " low is provided with " resource exhaustion " that may cause being used for little investigation, summed up summary on the identical server of little investigation with big investigation.Higher setting the to " h " may cause long being used to find the computing time of candidate.
As mentioned above, in one embodiment, make a summary and close record in device 230 processing queue, so that request is divided into groups in the following manner: not because of making that web feed request is processed and make the investigation exhaustion for looking after main investigation at special time.Then, the degree of depth that can be by control candidate search and the variable of storage control avoid resource exhaustion along with and the changes of properties of closing operation.In an alternate embodiment, make a summary and close device 230 and can check in the formation first simply, extract then and preceding " z " item of first coupling.Resource exhaustion will be avoided, but main investigation may be able to influence the summary of the investigation of negligible amounts.For example, may cause the obstruction of the summary speed of the investigation that influence is much smaller from a lot of requests of an investigation.But simpler method can provide configuration to go up less expense and complicacy.
Fig. 5 reads the process flow diagram 500 of the method for web feed request when showing and preparing for the summary that uses according to embodiments of the invention.At first, in step 510, engine adapter (reading) 310 is from making a summary and closing device 230 reception summaries and call.Then, in step 520, engine adapter (reading) 310 determines whether message relates to investigation submission incident or newly organize incident.
If web feed request message relates to investigation submission incident, then in step 530, engine adapter (reading) 310 reads to be used for the enquiry data set of new investigation response from database 120.Enquiry data set comprises about the structural information of investigation (for example, the group of relevant issues or project, the main scope or the theme that are made of subassembly).If web feed request message relates to new group of incident, then in step 535, engine adapter (reading) 310 is from the whole enquiry data set of database read.This data acquisition can comprise project for example, scope, the information that the response scale is relevant with other.
Then, for investigation submission incident,, read previous wrap-up state from database 120 in step 540.Previous wrap-up state can comprise project for example, scope, the information that responsive state is relevant with other.The previous original state of wrap-up state representative before the increment summary.
Then, for investigation submission incident, in step 550, find group at each line data, for example, according to these data, the respondent is the member of these data.Statistical treatment device 320 can be called, to determine which group is each single file data belong to.The group membership relation of investigation can be by based on the answer from investigation or precoding information, and described information is department or demographic information for example, is affixed on the respondent.For new group of incident,, find the new group of data line that will sum up summary in step 555.Statistical treatment device 320 can be called to determine which line data belongs to new group.Can calculate this group according to submitted investigation before creating this group at first.In one embodiment, group incident and investigation submission incident can be moved in turn establishments so that incorrect data are not searched into a matter, described search into a matter and group that it is affiliated processed at one time.The order of operation is inessential.
In step 560, read engine adapter (reading) 310 calls the calculating that statistical treatment device 320 is carried out the group of be affected (for new group, if incident is the incident of newly organizing).For example, the group that is applied to of structure, information, the data that will be made a summary and data can be passed to statistical treatment device 320.
In one embodiment, statistical treatment device 320 upgrades statistics or creates new statistics row according to new group according to new investigation response.For example, statistical treatment device 320 is carried out accrual accounting and is calculated, to determine the statistic behavior through adjusting according to the enquiry data that newly receives (for example, increment summary).Statistical treatment device 320 is also carried out the analysis to data, so that determine that for given respondent the group membership concerns.Allow read engine adapter 310 that group is loaded in the statistical abstract engine 3 00 by group analysis, statistical abstract engine 3 00 can be applicable to the specific one increment summary of taking turns.
In one embodiment, statistical treatment device 320 comprises the standard analysis device.The standard analysis device uses and for example to illustrate that logical statement carries out the semantic analysis to the investigation data stream.The standard analysis device is carried out the semantic analysis to data stream, to determine for example which group is which data belong to.Analyzer can be customized, so that implement to allow the operational character (operator) of faster and more effective enquiry data analysis.For example, the analyzer with dedicated operations symbol can be used Unix instrument lex and yacc generation, and is used C or the programming of C++ programming language.
Fig. 6 shows the process flow diagram 600 of the method that is used for the statistical abstract used according to embodiments of the invention.At first, in step 610, statistical treatment device 320 receives for example structure and other relevant informations of project, theme or scope, response scale, recompile bucket (recode bucket).Then, in step 620, statistical treatment device 320 receives the group that will be included in the summary.Then, in step 630, statistical treatment device 320 distributes Hash matrix data structure to come to keep original state for data.
In step 640, statistical treatment device 320 receives each associated row of data, and carries out the increment summary.That is the suitable row in the set of statistical treatment device 320 incremental data.To each line data, statistical treatment device 320 uses previous state and new data line at this row each group as its member.The Data Update Hash matrix that statistical treatment device 320 usefulness were upgraded.Then, in step 650, in case all line data are all processed, the wrap-up state of then upgrading is returned to engine adapter (writing) 330.
The Hash matrix structure that uses in summary is stored the statistics of investigation in the mode of saving storer.The Hash matrix structure is included in the Hash table on each of matrix.In one embodiment, the Hash matrix provides and has had the constant of minimum conflict and search the time.
In one embodiment, the statistics of Hash matrix storage investigation in " group " element of " project/scope " element of matrix and matrix.Group in the matrix is the group of related data row as its member, unless the Hash matrix is batch matrix (batch matrix) that comprises all groups.In one embodiment, the Hash matrix is the objects of statistics two-dimensional array that is coupled with three Hash table structures.The Hash table structure comprises the structure of project for example, structure and each scope of project or the structure of group of group.For example, the Hash table structure provides associative search in the Hash matrix on the right basis of project/group or scope/group.As a result, search (for example purpose in order to retrieve or to upgrade) of certain statistical carried out in the time of O (C).In one embodiment, make at possible minimized memory when keeping statistic behavior and be used for optimize storage and use.
In one embodiment, Hash table is the distribution configuration of 10 * N, and wherein N is the size of group element or the size of combination project and scope element.In typical case, the conflict in the Hash table is rare.Can use the higher Hash table configuration factor that distributes to reduce in some cases conflict.
For example, total storer utilization may be that 200 bytes * G * (I+D), wherein, G is the batch total number in the Hash matrix, and I is an item count, and D is scope (or group of project a) counting.For example, if G=10,000, and I+D=1000, the total memory utilization that then is used to keep wrap-up state comprises the space that Hash table is used a little more than the 2G byte.
In case summary is summed up and finished, the statistics that then is incremented is write back database 120.Control flow is passed back summary and is closed device 230.If more message still in message queue, is then made a summary and is closed device 230 and can repeat the statistical abstract process.The result that engine adapter (writing) 330 receiving and counting processors 320 are produced.The statistics that engine adapter (writing) 330 will upgrade is write database 120.Like this, database 120 is upgraded with by analysis investigation result and/or new group.
The statistics that specific project/group is right can comprise the response frequency of for example mean value, standard deviation, each response option, the recompile frequency of each recompile, actual N, the N and the stored non-response count of weighting.The statistics that particular range/group is right comprises the response frequency of for example mean value, standard deviation, each response option, the recompile frequency of each recompile, actual N, the N and the stored non-response count of weighting.
Can be each the problem counting statistics in the investigation.In addition, can be scope or theme counting statistics or the mark in the investigation.In one embodiment, calculating was done suppress data due to the respondent of not enough quantity before.In addition, ignore response outside effective answer scope.In one embodiment, be retrieved when data but not when statistics is calculated, round off.In one embodiment, the quilt of the space record in the database 120 is towards the total counting number of response but not at problem layer RESPONSE CALCULATION.In calculating, also can use reversing mark setting (for example, the response with from 1 to 5 effective value that changes can be registered as: 1=5,2=4,3=3,4=2 and 5=1).
Can use the inhibition of response to guarantee that each group has N response at least to problem or scope.Perhaps, according to the response of other groups, can suppress the result.Group and the relation that suppresses between the rule can be configured.For example, for particular problem or problem set, can give the response of a set of dispense minimum number or level.
At the collection approach that is used for determining statistics, create theme or scope result from problem level result.Come response frequency is sued for peace from the problem that comprises interested scope.The calculating and the RESPONSE CALCULATION that investigates a matter are similar, but respondent's quantity is by average at scope or theme mark.At first, response classification or bucket (for example approve of, neither, do not approve of, very, poor, fair, be, deny, or the like) by from being used for the problem interpolation of computer capacity.Then, add up, count from the response frequency bucket of scope and removed by the tale of all frequency buckets for number percent.Each response bucket can be assigned with one and be used for the value that for example mean value and standard deviation calculate.For mean value or average mark, range frequencies bucket value is multiplied by range frequencies bucket counting.Sum is removed by total respondent's quantity then.For example, mean value=((frequency bucket 1 * 1)+(frequency bucket 2 * 2)+(frequency bucket 3 * 3))/total respondent's quantity.Can be calculated as a lot of users of range response and to run through number average that the problem that is used for calculating this scope is made response.The respondent's of each problem quantity is added, and then, sum is divided by being used for the total quantity of problem of computer capacity.In one embodiment, the problem with zero respondent is included in the number of ranges of RESPONSE CALCULATION.By to non-response frequency bucket summation, can also calculate the counting of non-response.
In one embodiment, investigate a matter the statistics follow in the undefined rule of collection approach.Perhaps, can handle enquiry data with other statistical method.In one embodiment, if problem is included in the layout data file definition of group, then the problem result is generated at this group.
In the method for " by situation " that be used for definite statistics, each respondent who has answered at least one problem in the scope has been created the scope result.Respondent's range fraction is by average.At first, the problem result is handled at each respondent.Then, by to item response summation and divided by the quantity of the problem of in scope, being answered, for each respondent has created range fraction.Use following formula, can computational item mean value/average range mark from summed project bucket, for example: mean value=((frequency bucket 1 * 1)+(frequency bucket 2 * 2)+(frequency bucket 3 * 3))/run through total number of responses of project.By average respondent's range fraction, calculating can be done.
The method of a kind of " by project " uses collection approach described above to come the computational item result.Then, project result can be by on average, to create range fraction.In one embodiment, not satisfying the project that minimum N requires is left in the basket.In addition, range computation can be by based on the project result who rounds off.
Can also calculate main theme or scope.From subrange but not project makes up main scope.Subrange can be calculated based on the problem result.For example, use " by situation " or " by project " computing method, can use the subrange result to calculate main scope result.
When the respondent is added to group, can recomputate mean value and standard deviation.Can use following equation to come calculating mean value and standard deviation:
DX＝(X-S1)/N (1)；
DX2＝DX
2 (2)；
S2＝S2+(N×(N-1)×DX2)(3)；
S1＝S1+DX (4)；
SD＝SQR(S2/(N-1)) (5)，
Wherein, X is respondent's answer, and N is respondent's a quantity, and S1 represents mean value, and S2 indicates quadratic sum, and SD is a standard deviation, and SQR is a square root.Like this, utilize respondent's answer and respondent's quantity, can use previous mean value to calculate new mean value.Similarly, utilize respondent's answer and respondent's quantity, can use previous quadratic sum to determine the standard deviation of upgrading.
In one embodiment, can come weighting as a result, with reflection parent feature by the respondent.If weighted value is present in the tables of data that is used for a respondent, then this respondent's result is by this weighted value weighting.For weighted data, weighted value is added to data but not increases progressively the response bucket.Represent the statistics of respondent's quantity not to be weighted usually, but the people's of response true quantity is made in representative to a problem.
Statistical abstract engine 240-243 can comprise the routine that is used in a large number to add up with data.For example, engine 240-243 can comprise " acquisition data " routine that is used to return requested statistics.In one embodiment, the acquisition data routine had both contained inhibition and had also contained round-off result.In one embodiment, obtain data routine and get Several Parameters, comprise: statistics, the inhibition of expectation and the form that rounds off (for example quantity of radix point) that are used for requested group identifier, the identifier that is used for requested problem or scope, request type (for example problem or scope), expectation.For example, can use extra parameter to specify the condition of inhibition.The statistics of expectation can be for example frequency bucket (to the quantity of requested particular memory bucket response), number percent bucket (to the number percent of requested particular memory bucket response), non-response bucket (quantity of the non-response of bucket), mean value or average mark, standard deviation, to the quantity (actual N) of the significant response of requested problem or scope, and/or regardless of problem or scope, total quantity of the respondent of group (total N).The inhibition of expectation can comprise that for example unrestraint, a minimum N response, group suppress, custom suppresses (custom suppression) or other inhibition routine.
Obtain data routine and can return the statistics of expectation and the result of data search.Routine for example can return success (add up found and be not suppressed), be suppressed (add up found, but be suppressed owing to selected inhibition is regular) or invalid (do not find add up or do not have return data).
Statistical abstract engine 240-243 for example can also comprise " obtaining population (population) " routine.In one embodiment, obtain the population routine to counting, as the population size with the quantity of the clauses and subclauses of the matches criteria of group.Perhaps, all records in set of records ends can be returned, so that come self-recording information to be printed.
Another routine is " acquisition written comment " routine.Obtain the written comment routine from designated group written comment document retrieval written comment.For example, written comment can be returned by problem and/or by the topic in the problem.Except the comment of reality, other statistics can also be provided,, provide quantity the respondent of the comment of this problem for example from the quantity of the comment of the problem of this group, the number percent of the comment in the quantity of the comment in the topic of written comment problem and the topic of written comment problem.
In an alternate embodiment, system 100 can comprise reporting system (not drawn).For example, reporting system can play the effect that definition is used to analyze the batch processor of the business rule of analyzing with investigation statistics.In one embodiment, reporting system comprises single statistical treatment device, and this processor is carried out for example accrual accounting and handled.In one embodiment, message is not re-used as in Fig. 2 or Fig. 3 or and closes.Data and group are by retrieval in the statistical treatment device of reporting system and sum up summary.For example, statistics can be shown, print and/be stored in the database 120.For example, can the operation report system define and analysis bank, scope project, response, data field, and/or report data.Reporting system is new database 120 more, is used for and investigation and analysis system 200 uses together.
Fig. 7 shows according to embodiments of the invention, is used to produce the report layout system 700 that shows the investigation and analysis data document.In one embodiment, for example generation of the report layout of the portable document files of Adobe (PDF) form, Microsoft text formatting or Microsoft spreadsheet format has been simplified by system 700.In one embodiment, system 700 comprises font module 710, color module 720 and PDF module 730.In one embodiment, use together in system's 700 quilts and PDFlib and PDFGen storehouse, to produce the report of PDF, comprises text and/or figure.In addition, can use application program, for example Excel Writer
TM(electrical form Write) comes to produce MicrosoftExcel (Microsoft's electrical form) document on WWW.
The font module 710 auxiliary fonts of the report of pdf document of for example reporting are selected and generation.In one embodiment, font module 710 is common application DLL (dynamic link library) (API).Font module 710 has encapsulated the font of using in producing PDF or other document format files.The font that font module 710 allows to be independent of the work of specific character collection is selected.For example, print the japanese character character if just using based on the font of latin text, then font module 710 substitutes appropriate font.
In order to dispose the font that is used to report, font module 710 can be used the attribute that for example embeds (Embeded), coding (Encoding), fontname (FontName), font ratio (FontScale), GetLangEnum, overstriking (IsBold), tilt (IsItalic), underscore (IsUnderlined), unified character code (IsUnicode), language and size.Font module can also for example be used the method for duplicating (Copy), GetLanguageList, SetFontByName, SetFontCustom and SetStyle.
For example, color module 720 also may be implemented as public API.In one embodiment, when producing the document of pdf document for example, color module 720 has been simplified to the object branch and has been mixed colours.For example, color module 720 allows to use the RGB tlv triple or creates color from " designated color (namedcolors) " that enumerate that set in advance.Produce for color, color module 720 can be used for example attribute of color (for example red, green and blue) and name.Color module 720 can also be used and duplicate, the method for SetByName and SetByRGB.
In order to generate the report file layout, PDF module 730 can be used for example attribute of founder (Creator), current page (CurrentPage), document status (DocumentStatus) and m PDF.PDF module 730 can also be used for example AddBookMark (interpolation book label), AddImage (interpolation image), AddImageInBox (in frame, adding image), AddWebLinkBox (adding the network linking frame), BeginPage (beginning page or leaf), CloseFile (close file), Copy (duplicating), CurHGraphPos (the present level position of graphic cursor), CurHPos (the present level position of text cursor), CurPageHeight, CurPageWidth, CurVGraphPos, (graphic cursor upright position), CurVPos (text cursor upright position), DrawShape is (for example excellent, line, frame, circle, polygon), EndPage (last page or leaf), GetActualX/Y (obtaining actual X/Y), GetMatchingPenColor (color that obtains mating), HTMLColorToRGB (the HTML color is to RGB), MoveHPos (mobile and horizontal position), MoveVPos (mobile upright position), OpenFile (opening file), PlacePDI (placing PDI), PrintText (print text), PrintTextInBox (printing the text in the frame), PrintTextInBoxHTML (printing frame Chinese version HTML), the method of ScalePage (yardstick page or leaf) and SetCursomPageSize (client's page or leaf size is set).
Therefore, some embodiments of the present invention have been showed the real-time electronic system that investigates.Some embodiment allow the user to create investigation and collect data.Can define for example problem, group and scope (theme).Engine can deal with data.Then, produce report from processed data.
Fig. 8 shows the process flow diagram 800 that is used for e-survey management of using according to embodiments of the invention and the method for analyzing.At first, in step 810, editor's investigation participator's tabulation.For example, can tabulate by editing e-mail, be used for using in management survey.Then, in step 820, investigation is distributed to the investigation participator in the tabulation.For example, use the link and the password of investigation webpage, each participator on the email list is issued in investigation by Email.
Then, in step 830, the investigation participator is connected to the investigation and management system, and finishes investigation.For example, investigation participator accessed web page participates in investigation.Participator's identity can be certified.Then, in step 840, participator's response is hunted down.For example, based on network form is caught participator's response.
In step 850, sign except respondent's identifier and password information are peeled off from enquiry data.Then, in step 860, enquiry data is sent to be handled and reporting system.In an alternate embodiment, the papery investigation can be transfused to data file and be used for handling.At last, in step 870, can produce the report of survey data.
Investigate And Report shows can be translated into different language.In one embodiment, when investigation is created, can submit the translation that investigates a matter to, and response and final output are translated correspondingly.
In one embodiment, the keeper can select different rules for handling enquiry data.Management can be provided with different responses, response scale and/or with reference to project for investigation.The type of statistical report and pattern can be selected, and send into reporting system 700.Group can be by the sets definition according to certain value of selecteed problem.Can also use the grammer interpreter to select group.Interpreter can use the statistical operation of definition of data group to accord with construction data case statement.
In an alternate embodiment, can use sampling instrument to come sampling survey data and statistics.Sampling instrument can be carried out statistical computation to enquiry data.Sampling instrument can use the investigation response data to carry out random statistical and calculate and layering statistical computation (for example, data are divided into group).
Like this, some embodiments of the present invention provide the system and method that is used to analyse and investigate with statistics.Some embodiment use distributed computing technology to carry out for example general purpose and the statistical computation of the enquiry data of feedback data at many levels.Some system provides scalable system, and scalable system can add or cut down hardware, does not influence software or function to hold the investigation load.Some embodiment comprise and the matrix structure of Hash table coupling, are used to run through whole investigation and efficiently handle investigation statistics.In certain embodiments, can use based on the standard analysis device of outside definable semantic rules and for example determine which group which investigation response belongs to.In one embodiment, the user can define standard statement and real-time retrieval result.Some embodiment are multiplexing investigation web feed request between a plurality of summary servers, to allow to run through the peak performance of a plurality of investigation.A plurality of investigation and/or a plurality of database can be received.Therefore, some embodiment provide the real-time processing to enquiry data, and allow the immediate feedback during fact-finding process.
Though the present invention has been described with reference to some embodiment,, it will be appreciated by those skilled in the art that and can make various changes, and, not departing from scope of the present invention, equivalent can be replaced.In addition, according to instruction of the present invention, do not depart from scope of the present invention and can make a lot of modifications, to adapt to particular case or material.Therefore, expectation the present invention is not limited to the disclosed embodiments, and the present invention will comprise all embodiment that fall in the claims scope.
Claims (21)
1. disposal system that is used to handle enquiry data, described system comprises:
The read engine adapter is used to read the starting point that enquiry data and mark are used to handle enquiry data;
The survey processor is used to use semantic analysis to handle described enquiry data, and described survey processor is at least one the generation statistics in investigation, problem and the scope according to described enquiry data; With
Write the engine adapter, be used to export the investigation statistics of renewal.
2. the system as claimed in claim 1, wherein, described read engine adapter reads previous wrap-up state from database.
3. the system as claimed in claim 1, wherein, described survey processor comprises and is used for the standard analysis device enquiry data sorted out according to the definable semantic rules.
4. the system as claimed in claim 1, wherein, the described enquiry data of described survey processor flag is as its member's report group.
5. the system as claimed in claim 1 also comprises the Hash matrix of the statistics that is used to store investigation.
6. real-time investigation and analysis system, described system comprises:
Run time engine is used to receive enquiry data;
At least one summary also closes device, is used for the Classification Count data, and described enquiry data is routed at least one statistical abstract engine, and described at least one summary also closes device and classifies and the described enquiry data of route according at least one standard; With
At least one statistical abstract engine is used for analyzing in real time described enquiry data, with in executive summary and the statistical study at least one.
7. system as claimed in claim 6, wherein, described at least one summary also closes device from a plurality of sources reception enquiry data streams.
8. system as claimed in claim 6, wherein, described enquiry data comprises investigates at least one that finish in data and the new survey report group data.
9. system as claimed in claim 6, wherein, described at least one standard also comprises the described summary of permission and closes device according at least one standard of the enquiry data that enters being divided into groups for the order of handling optimum.
10. system as claimed in claim 6, wherein, described summary also closes device the history list of a plurality of enquiry datas with recently processed investigation is compared, so that described a plurality of enquiry datas are routed to described at least one statistical abstract engine.
11. system as claimed in claim 6, wherein, described at least one statistical abstract engine reads previous wrap-up state from database, so that the starting point of using described enquiry data mark accrual accounting to calculate.
12. system as claimed in claim 6 also comprises report engine, is used to produce the new group that is used for survey report.
13. system as claimed in claim 6 also comprises database, is used for storing at least one of enquiry data and the definition of survey report group.
14. system as claimed in claim 6 also comprises the user interface that is used to import enquiry data.
15. system as claimed in claim 14, wherein, described user interface also comprises based on network user interface.
16. a method that is used for handling in real time enquiry data, described method comprises:
Reception searches into a matter;
Determine the described route that searches into a matter according at least one standard;
Described searching into a matter is routed to the device of investigating;
Handle described searching into a matter in real time at the described device of investigating; With
The real-time feedback that searches into a matter from described is provided.
17. method as claimed in claim 16 also comprises and determines whether the incident that enters comprises enquiry data or new investigation group.
18. method as claimed in claim 16, wherein, described treatment step also comprises enquiry data is carried out statistical treatment.
19. method as claimed in claim 16, wherein, described treatment step also comprises creates new investigation group.
20. method as claimed in claim 16, wherein, described investigation step also comprises according to investigation, problem and wherein at least one of theme analyses and investigates response.
21. method as claimed in claim 16 also comprises according to described processed searching into a matter and upgrades stored data.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US47122303P | 2003-05-16 | 2003-05-16 | |
| US60/471,223 | 2003-05-16 | ||
| US10/844,067 | 2004-05-11 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1973281A true CN1973281A (en) | 2007-05-30 |
Family
ID=38113176
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200480013206 Pending CN1973281A (en) | 2003-05-16 | 2004-05-14 | Method and apparatus for survey processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1973281A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010075671A1 (en) * | 2008-12-31 | 2010-07-08 | 北京思在信息技术有限责任公司 | Method and apparatus for updating abstract structure |
| CN112825135A (en) * | 2019-11-20 | 2021-05-21 | 株式会社理光 | Display device, display method, and medium |
-
2004
- 2004-05-14 CN CN 200480013206 patent/CN1973281A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010075671A1 (en) * | 2008-12-31 | 2010-07-08 | 北京思在信息技术有限责任公司 | Method and apparatus for updating abstract structure |
| CN112825135A (en) * | 2019-11-20 | 2021-05-21 | 株式会社理光 | Display device, display method, and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7418496B2 (en) | Method and apparatus for survey processing | |
| US10592705B2 (en) | System and method for network user interface report formatting | |
| US6826724B1 (en) | Document processor, document classification device, document processing method, document classification method, and computer-readable recording medium for recording programs for executing the methods on a computer | |
| KR101067398B1 (en) | Method and computer-readable medium for importing and exporting hierarchically structured data | |
| US9922096B2 (en) | Automated presentation of information using infographics | |
| US7664732B2 (en) | Method of managing websites registered in search engine and a system thereof | |
| US6704723B1 (en) | Method and system for providing business intelligence information over a computer network via extensible markup language | |
| CN108776671A (en) | A kind of network public sentiment monitoring system and method | |
| CN111708774B (en) | Industry analytic system based on big data | |
| US6754654B1 (en) | System and method for extracting knowledge from documents | |
| US11763180B2 (en) | Unsupervised competition-based encoding | |
| CN1973281A (en) | Method and apparatus for survey processing | |
| CN116089490A (en) | Data analysis method, device, terminal and storage medium | |
| CN110716994B (en) | Retrieval method and device supporting heterogeneous geographic data resource retrieval | |
| US9183317B1 (en) | System and method for exporting report results from a reporting system | |
| CN115689463A (en) | Enterprise standing book database management system in rare earth industry | |
| Rong et al. | Direct out-of-memory distributed parallel frequent pattern mining | |
| JP2021107966A (en) | Information processing program, information processing method, and information processing apparatus | |
| CN114996320A (en) | Family activity recommendation method and system and electronic equipment | |
| KR20020069762A (en) | System and method for searching an appointed web site | |
| Eichstädt | Internet webcasting: generating and matching profiles | |
| CN110765129A (en) | High-performance online expense settlement statistical method and device | |
| CN115982293A (en) | Information technology service system based on Internet of things | |
| CA2400489C (en) | Information delivery system and method | |
| CN118628151A (en) | Policy information screening and customer matching equipment, system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |