Method for quantitatively evaluating resident travel questionnaire quality

Abstract

The invention belongs to the technical field of information survey processing, and particularly relates to a method for quantitatively evaluating resident travel questionnaire quality, which comprises the following steps of: S100, marking a questionnaire based on the consistency of a questionnaire user address and a questionnaire position; s200, marking the questionnaire based on the integrity of the questionnaire data; s300, comprehensively analyzing the rationality of the questionnaire user information, the family vehicle information and the personal basic information, and marking questionnaires; s400, a questionnaire is marked based on the rationality of the personal travel information of the previous day; s500, analyzing the similarity of different questionnaire travel destinations of the same questionnaire based on an editing distance method; and S600, based on a weighted average method, calculating the scores of the investigators. The questionnaire quality can be quickly audited, and the quality of the questionnaire is evaluated based on the questionnaire audit result, so that the questionnaire audit efficiency is greatly improved, and the manual audit workload is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of information survey processing, and in particular to a method for quantitatively evaluating the quality of residents' travel survey questionnaires. Background Art

[0002] Resident travel surveys are the main means of understanding the travel characteristics of residents, and can provide a basis for decision-making for government departments and research institutions. Therefore, it is crucial to ensure the quality of questionnaires. Common survey quality control measures are mainly formulated from four aspects: survey organization, survey team building, survey methods, and questionnaire quality inspection. Among them, questionnaire quality inspection is the last key link in survey quality control. The traditional questionnaire quality inspection method usually organizes auditors to manually check household information, family vehicle information, personal basic information and travel information data from the previous day. There are shortcomings such as time-consuming and labor-intensive questionnaire review, and the inability to timely feedback on questionnaire quality and investigator quality, which makes it difficult to repair problematic questionnaires in a timely manner and it is difficult to improve the quality of investigators in a targeted manner. With the promotion of electronic resident travel survey procedures, questionnaire data can be uploaded to the system background in real time after the household survey is completed, and the efficiency of survey data aggregation has been greatly improved. The traditional questionnaire quality inspection method can no longer meet the actual needs.

[0003] To this end, the present invention designs a method for quantitatively evaluating the quality of residents' travel questionnaires, which can quickly review the quality of questionnaires and evaluate the quality of investigators based on the questionnaire review results, thereby greatly improving the efficiency of questionnaire review and reducing the workload of manual review. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a method for quantitatively evaluating the quality of residents' travel questionnaires.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for quantitatively evaluating the quality of a resident travel questionnaire includes the following steps:

[0007] S100: Marking the questionnaire based on the consistency between the household address and the investigator's location;

[0008] S200: Marking questionnaires based on completeness of questionnaire data;

[0009] S300: Comprehensively analyze the rationality marking questionnaire of household information, family vehicle information, and personal basic information;

[0010] S400: Rationality marking questionnaire based on personal travel information the day before;

[0011] S500: Analyze the similarity of travel destinations of different questionnaires of the same investigator based on the edit distance method;

[0012] S600: Calculate the investigator score based on the weighted average method.

[0013] Furthermore, the S100 specifically includes household information, family vehicle information, basic personal information, personal travel information the day before, and investigator's household location information;

[0014] The coordinates of the household address of the questionnaire and the location of the investigator entering the household are converted into Gauss-Kruger projection coordinates, the coordinates of the location of the investigator entering the household on the day of the questionnaire survey are screened out, and the distance between the household address of the questionnaire and the coordinates of the location of the investigator entering the household is calculated. The calculation formula is:

[0015]

[0016] In the formula, distance represents the distance between two projection coordinates, (x 1 ,y 1 )、(x 2 ,y 2 ) represents two projection coordinates; select the minimum distance from all distance data min , if 300 <distance min ≤1000, mark the questionnaire as "Position Pending"; if distance min >1000, mark the questionnaire as “failed”.

[0017] Furthermore, the household information includes the household address including longitude and latitude, household type, number of permanent residents in the household, number of preschool children, annual family income range, housing type, household type, number of cars and electric bicycles owned by the family; family vehicle information includes vehicle type, power type, average annual mileage, main purpose of the vehicle, daily parking conditions, and future car purchase intentions; personal basic information includes household registration status, gender, age, occupation, industry, education level, whether they have a driver's license, and whether they drive on a regular basis; personal travel information the day before includes the starting and ending points, start and end times, travel purposes, and modes of transportation used for each trip of household members on the previous working day; the investigator's household location information is the time and location coordinate information when the investigator visits the household to investigate the data.

[0018] Furthermore, in said S200, the specific steps are: repeatedly judging whether the data in the questionnaire is complete, marking the questionnaires that lack household information, vehicle information or personal basic information as "completeness pending"; marking the questionnaires whose household address coordinates are not within the preset survey range as "completeness pending"; marking the questionnaires that lack the travel information of the previous day as "failed".

[0019] Furthermore, in the S300, the specific steps are: counting the sample size of cars and electric bicycles in the family vehicle information of the questionnaire, marking the questionnaires whose number of cars owned by the family in the household information is inconsistent with the sample size of cars in the family vehicle information as "information rationality pending", marking the questionnaires whose number of electric bicycles owned by the family in the household information is inconsistent with the sample size of electric bicycles in the family vehicle information as "information rationality pending";

[0020] The number of individual samples in the basic personal information of the questionnaire was counted, and the questionnaires with a difference of 1 between the number of permanent residents in the household information and the number of individual samples in the basic personal information were marked as "information rationality pending", and the questionnaires with a difference of more than 1 between the number of permanent residents in the household information and the number of individual samples in the basic personal information were marked as "failed";

[0021] Extract the sample of retired personnel from the personal basic information in the questionnaire. If the male retiree is younger than 50 years old or the female retiree is younger than 45 years old, mark the questionnaire as "information rationality pending";

[0022] Individual samples with father-son, father-daughter, mother-son, or mother-daughter relationships in their basic personal information were extracted. If the age difference between father and son or father and daughter was less than 22 years old, the questionnaire was marked as "information rationality pending"; if the age difference between mother and son or mother and daughter was less than 20 years old, the questionnaire was marked as "information rationality pending".

[0023] Further, in the S400, the specific steps are: extracting the mode of transportation used by the individual on the previous day, marking the questionnaires in which the individual used an electric bicycle on the previous day and the number of electric bicycles owned by the family in the household information is 0 as "travel rationality pending", marking the questionnaires in which the individual used a private car on the previous day and the number of private cars owned by the family in the household information is 0 as "travel rationality pending";

[0024] Traverse the individual's travel information of the previous day, extract the starting point coordinates, end point coordinates, departure time, arrival time, and travel mode of each trip of the individual on the previous day, and calculate the travel duration obtained from the survey; use the starting point coordinates, end point coordinates, and travel mode as input, and obtain the total distance, time, cost, and driving path data of the actual shortest time path between the start and end points through the route planning interface of different travel modes on the Internet map; calculate the relative difference between the survey travel time and the actual travel time, and mark the questionnaires with a relative difference exceeding the threshold (30%) as "travel rationality pending".

[0025] Furthermore, in S500, the specific steps are: traverse the investigators, extract the addresses of the personal travel destinations the day before in all the questionnaires of the investigators, eliminate the destinations that are the same as the household addresses, and use the edit distance algorithm to calculate the similarity between different addresses. The edit distance algorithm is:

[0026] (1) Define an array; for two address strings A and B of length m and n respectively, define a two-dimensional array D, where D[i][j] represents the minimum edit distance for converting the first i characters of string A to the first j characters of string B;

[0027] (2) Initialize boundary value;

[0028] When i=0 and j=0, let D[0][0]=0, indicating that the edit distance between two empty strings is 0;

[0029] When i>0 and j=0, let D[i][0]=i, which means that the minimum edit distance to change the first i characters of string A into an empty string is i, corresponding to the complete deletion operation;

[0030] When i=0 and j>0, let D[0][j]=j, which means the minimum edit distance to convert an empty string into the first j characters of string B is j, corresponding to all insertion operations;

[0031] (3) Edit distance recursion;

[0032] When i>0 and j>0, if A[i]=B[j], then D[i][j]=D[i-1][j-1], that is, the current character matches and the edit distance is equal to the previous one; if A[i] is not equal to B[j], then D[i][j]=min{D[i-1][j]+1,D[i][j-1]+1,D[i-1][j-1]+1}, which correspond to deletion, insertion and replacement operations respectively, and the minimum value of the three is taken as the current edit distance;

[0033] (4) Output the edit distance; D[m][n] is the edit distance of converting string A to string B;

[0034] Traverse the address string A of the investigator, and regard other addresses B whose edit distance to the length of A is less than the threshold (20%) and whose actual distance is less than 300m as similar addresses to address A. Count the proportion of addresses similar to address A to the total number of addresses. If the proportion is greater than the threshold (10%), mark the questionnaire containing address A and its similar addresses as "address similarity pending";

[0035] Traverse the investigators and count the proportion of male individual samples in all questionnaires of the investigators. If the male proportion exceeds the threshold (greater than 60% or less than 40%), the investigator is marked as "abnormal".

[0036] Furthermore, in S500, the specific steps are: calculating the investigator score according to the investigator's questionnaire marking; setting the investigator's initial score, such as 10 points; setting corresponding deduction weights for different questionnaire marks, such as setting the deduction weights w for the marks of location pending, completeness pending, information rationality pending, travel rationality pending, and address identical pending. 1 、w 2 、w 3 、w 4 、w 5 They are set to -3, -1, -1, -2, and -0.5 respectively, and then the weighted average method is used to calculate the investigator's deduction value. The difference between the investigator's initial score and the deduction value is taken as the final score. The calculation formula is:

[0037]

[0038] In the formula, totalscore represents the investigator's initial score, N represents the number of questionnaires of the investigator, and x i,j Indicates whether questionnaire i has the jth mark, w j Indicates the weight of the jth mark; the scores are sorted from large to small and displayed as the basis for evaluating investigators.

[0039] The above technical solution of the present invention has the following beneficial technical effects:

[0040] The present invention can quickly review the quality of questionnaires and evaluate the quality of investigators based on the questionnaire review results, thereby greatly improving the efficiency of questionnaire review and reducing the workload of manual review. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a method for quantitatively evaluating the quality of residents' travel questionnaires proposed by the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0043] like Figure 1 As shown, the present invention provides a method for quantitatively evaluating the quality of residents' travel questionnaires, comprising the following steps:

[0044] Step 1: Label the questionnaire based on the consistency between the household address and the investigator location.

[0045] The questionnaire data collected by the residents' travel survey program are obtained, including household information, family vehicle information, personal basic information, personal travel information the day before, and investigators' household location information. Among them, household information includes household address (including latitude and longitude), household type, number of permanent residents in the household, number of preschool children, annual family income range, housing type, household type, number of cars and electric bicycles owned by the family, etc.; family vehicle information includes vehicle type, power type, annual average mileage, main purpose of the vehicle, daily parking conditions, future car purchase intentions, etc.; personal basic information includes household registration status, gender, age, occupation, industry, education, whether there is a driver's license, whether to drive, etc.; personal travel information the day before includes the starting and ending points, starting and ending times, travel purposes, and transportation methods used for each trip of household members on the previous working day; investigators' household location information is the time and location coordinates of the investigators when they visit the households to investigate data.

[0046] The coordinates of the household address of the questionnaire and the location of the investigator entering the household are converted into Gauss-Kruger projection coordinates, the coordinates of the location of the investigator entering the household on the day of the questionnaire survey are screened out, and the distance between the household address of the questionnaire and the coordinates of the location of the investigator entering the household is calculated. The calculation formula is:

[0047]

[0048] In the formula, distance represents the distance between two projection coordinates, (x 1 ,y 1 )、(x 2 ,y 2 ) represents two projection coordinates. Select the minimum distance from all distance data min , if 300 <distance min ≤1000, mark the questionnaire as "Position Pending"; if distance min >1000, mark the questionnaire as “failed”.

[0049] Step 2: Mark the questionnaire based on the completeness of the questionnaire data.

[0050] Repeatedly determine whether the data in the questionnaire is complete, and mark the questionnaires that lack household information, vehicle information or personal basic information as "completeness pending"; mark the questionnaires whose household address coordinates are not within the preset survey range as "completeness pending"; mark the questionnaires that lack travel information from the previous day as "failed".

[0051] Step 3: Comprehensively analyze the household information, family vehicle information, and personal basic information in the questionnaire to mark the questionnaire for rationality.

[0052] The sample size of cars and electric bicycles in the household vehicle information of the statistical questionnaire was calculated. The questionnaires whose number of cars owned by the household in the household information was inconsistent with the sample size of cars in the household vehicle information were marked as "information rationality pending". The questionnaires whose number of electric bicycles owned by the household in the household information was inconsistent with the sample size of electric bicycles in the household vehicle information were marked as "information rationality pending".

[0053] The number of individual samples in the personal basic information of the questionnaire is counted, and the questionnaires with a difference of 1 between the number of permanent residents in the household information and the number of individual samples in the personal basic information are marked as "information rationality pending", and the questionnaires with a difference of more than 1 between the number of permanent residents in the household information and the number of individual samples in the personal basic information are marked as "failed".

[0054] Extract the sample of retired personnel from the personal basic information in the questionnaire. If the male retiree is under 50 years old or the female retiree is under 45 years old, mark the questionnaire as "information rationality pending".

[0055] Individual samples with father-son, father-daughter, mother-son, or mother-daughter relationships in their basic personal information were extracted. If the age difference between father and son or father and daughter was less than 22 years old, the questionnaire was marked as "information rationality pending"; if the age difference between mother and son or mother and daughter was less than 20 years old, the questionnaire was marked as "information rationality pending".

[0056] Step 4: Mark the questionnaire based on the rationality of the individual’s travel information from the previous day.

[0057] The mode of transportation used by the individual in the previous day’s travel was extracted, and questionnaires where the individual used an electric bicycle in the previous day’s travel but the number of electric bicycles owned by the household in the household information was 0 were marked as “trip rationality pending”; questionnaires where the individual used a private car in the previous day’s travel but the number of private cars owned by the household in the household information was 0 were marked as “trip rationality pending”.

[0058] Traverse the individual's travel information of the previous day, extract the starting point coordinates, end point coordinates, departure time, arrival time, and travel mode of each trip of the individual on the previous day, and calculate the travel duration obtained from the survey; use the starting point coordinates, end point coordinates, and travel mode as input, and obtain the total distance, time, cost, and driving path data of the actual shortest time path between the start and end points through the route planning interface of different travel modes on the Internet map; calculate the relative difference between the survey travel time and the actual travel time, and mark the questionnaires with a relative difference exceeding the threshold (30%) as "travel rationality pending".

[0059] Step 5: Analyze the similarity of travel destinations of the same investigator in different questionnaires based on the edit distance method.

[0060] 5.1 Traverse the investigators, extract the addresses of the personal travel destinations the day before in all the questionnaires of the investigators, remove the destinations that are the same as the household addresses, and use the edit distance algorithm to calculate the similarity between different addresses. The edit distance algorithm is:

[0061] (1) Define an array. For two address strings A and B with lengths of m and n respectively, define a two-dimensional array D, where D[i][j] represents the minimum edit distance for converting the first i characters of string A to the first j characters of string B.

[0062] (2) Initialize the boundary value.

[0063] When i=0 and j=0, let D[0][0]=0, indicating that the edit distance between two empty strings is 0;

[0064] When i>0 and j=0, let D[i][0]=i, which means that the minimum edit distance to change the first i characters of string A into an empty string is i, corresponding to the complete deletion operation;

[0065] When i=0 and j>0, let D[0][j]=j, which means that the minimum edit distance to convert an empty string into the first j characters of string B is j, corresponding to all insertion operations.

[0066] (3) Edit distance recursion.

[0067] When i>0 and j>0, if A[i]=B[j], then D[i][j]=D[i-1][j-1], that is, the current character matches and the edit distance is the same as before; if A[i] is not equal to B[j], then D[i][j]=min{D[i-1][j]+1,D[i][j-1]+1,D[i-1][j-1]+1}, which correspond to deletion, insertion and replacement operations respectively, and the minimum value of the three is taken as the current edit distance.

[0068] (4) Output the edit distance. D[m][n] is the edit distance of converting string A to string B.

[0069] Traverse the investigator's address string A, and regard other addresses B whose ratio of edit distance to A length is less than the threshold (20%) and whose actual distance is less than 300m as similar addresses to address A. Count the proportion of addresses similar to address A to the total number of addresses. If the proportion is greater than the threshold (10%), the questionnaires containing address A and its similar addresses will be marked as "address similarity pending".

[0070] 5.2 Traverse the investigators and count the proportion of male individual samples in all questionnaires of the investigators. If the male proportion exceeds the threshold (greater than 60% or less than 40%), the investigator is marked as "abnormal".

[0071] Step 6: Calculate the investigator score based on the weighted average method.

[0072] Calculate the investigator score based on the investigator's questionnaire marks. Set the investigator's initial score, such as 10 points; set corresponding deduction weights for different questionnaire marks, such as setting the deduction weights for location pending, completeness pending, information rationality pending, travel rationality pending, and address duplication pending 1 、w 2 、w 3 、w 4 、w 5 They are set to -3, -1, -1, -2, and -0.5 respectively, and then the weighted average method is used to calculate the investigator's deduction value. The difference between the investigator's initial score and the deduction value is taken as the final score. The calculation formula is:

[0073]

[0074] In the formula, totalscore represents the investigator's initial score, N represents the number of questionnaires of the investigator, and x i,j Indicates whether questionnaire i has the jth mark, w j Indicates the weight of the jth mark. The scores are sorted from large to small and displayed as the basis for evaluating investigators.

[0075] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention is described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for quantitatively evaluating the quality of residents' travel questionnaires, characterized in that: The method comprises the following steps: S100: Marking the questionnaire based on the consistency between the household address and the investigator's location; S200: Marking questionnaire based on completeness of questionnaire data; S300: Comprehensively analyze the rationality marking questionnaire of household information, family vehicle information, and personal basic information; S400: Questionnaire for rationality marking based on personal travel information from the previous day; S500: Analyze the similarity of travel destinations of different questionnaires of the same investigator based on the edit distance method; S600: Calculate the investigator score based on the weighted average method.

2. A method for quantitatively evaluating the quality of residents' travel questionnaires according to claim 1, characterized in that: The S100 specifically includes household information, family vehicle information, basic personal information, personal travel information the day before, and investigator's household location information; The coordinates of the household address of the questionnaire and the location of the investigator entering the household are converted into Gauss-Kruger projection coordinates, the coordinates of the location of the investigator entering the household on the day of the questionnaire survey are screened out, and the distance between the household address of the questionnaire and the coordinates of the location of the investigator entering the household is calculated. The calculation formula is: In the formula, distance represents the distance between two projection coordinates, (x1, y1) and (x2, y2) represent two projection coordinates; select the minimum distance distance from all distance data min , if 300 <distance min ≤1000, mark the questionnaire as "Position Pending"; if distance min >1000, mark the questionnaire as "failed".

3. A method for quantitatively evaluating the quality of residents' travel questionnaires according to claim 2, characterized in that: The household information includes the household address including longitude and latitude, household type, number of permanent residents in the household, number of preschool children, annual household income range, housing type, household layout, number of cars and electric bicycles owned by the household; household vehicle information includes vehicle type, power type, average annual mileage, main purpose of the vehicle, daily parking conditions, and future vehicle purchase intentions; personal basic information includes household registration status, gender, age, occupation, industry, education level, whether they have a driver's license, and whether they drive regularly; personal travel information on the previous day includes the starting and ending points, start and end times, travel purposes, and modes of transportation used for each trip of household members on the previous working day; the household location information of the investigator is the time and location coordinate information when the investigator visits the household to investigate the data.

4. A method for quantitatively evaluating the quality of residents' travel questionnaires according to claim 1, characterized in that: In the S200, the specific steps are: repeatedly judging whether the data in the questionnaire are complete, marking the questionnaires that lack household information, vehicle information or personal basic information as "completeness pending"; marking the questionnaires whose household address coordinates are not within the preset survey range as "completeness pending"; marking the questionnaires that lack the travel information of the previous day as "failed".

5. The method for quantitatively evaluating the quality of residents' travel questionnaire according to claim 1, characterized in that: In the S300, the specific steps are: counting the sample size of cars and electric bicycles in the family vehicle information of the questionnaire, marking the questionnaires whose number of cars owned by the family in the household information is inconsistent with the sample size of cars in the family vehicle information as "information rationality pending", and marking the questionnaires whose number of electric bicycles owned by the family in the household information is inconsistent with the sample size of electric bicycles in the family vehicle information as "information rationality pending"; The number of individual samples in the basic personal information of the questionnaire was counted, and the questionnaires with a difference of 1 between the number of permanent residents in the household information and the number of individual samples in the basic personal information were marked as "information rationality pending", and the questionnaires with a difference of more than 1 between the number of permanent residents in the household information and the number of individual samples in the basic personal information were marked as "failed"; Extract the sample of retired personnel from the personal basic information in the questionnaire. If the male retired personnel are under 50 years old or the female retired personnel are under 45 years old, mark the questionnaire as "information rationality pending"; Extract personal samples with father-son, father-daughter, mother-son, and mother-daughter relationships in personal basic information. If the age difference between father and son or father and daughter is less than 22 years old, mark the questionnaire as "information rationality pending"; if the age difference between mother and son or mother and daughter is less than 20 years old, mark the questionnaire as "information rationality pending".

6. A method for quantitatively evaluating the quality of residents' travel questionnaires according to claim 1, characterized in that: In the S400, the specific steps are: extracting the mode of transportation used by the individual on the previous day, marking the questionnaires where the individual used an electric bicycle on the previous day and the number of electric bicycles owned by the household in the household information is 0 as "travel rationality pending", marking the questionnaires where the individual used a private car on the previous day and the number of private cars owned by the household in the household information is 0 as "travel rationality pending"; Traverse the individual's travel information of the previous day, extract the starting point coordinates, end point coordinates, departure time, arrival time, and travel mode of each individual's trip on the previous day, and calculate the travel time obtained from the survey; use the starting point coordinates, end point coordinates, and travel mode as input, and obtain the total distance, time, cost, and driving path data of the actual shortest time path between the starting and ending points through the route planning interface of different travel modes on the Internet map; calculate the relative difference between the survey travel time and the actual travel time, and mark the questionnaires with a relative difference exceeding the threshold (30%) as "travel rationality pending".

7. A method for quantitatively evaluating the quality of residents' travel questionnaires according to claim 1, characterized in that: In S500, the specific steps are: traverse the investigators, extract the addresses of the personal travel destinations the day before in all the questionnaires of the investigators, eliminate the destinations that are the same as the household addresses, and use the edit distance algorithm to calculate the similarity between different addresses. The edit distance algorithm is: (1) Define an array; for two address strings A and B of length m and n respectively, define a two-dimensional array D, where D[i][j] represents the minimum edit distance for converting the first i characters of string A to the first j characters of string B; (2) Initialize boundary value; When i=0 and j=0, let D[0][0]=0, indicating that the edit distance between two empty strings is 0; When i>0 and j=0, let D[i][0]=i, which means the minimum edit distance to change the first i characters of string A into an empty string is i, corresponding to the complete deletion operation; When i=0 and j>0, let D[0][j]=j, which means the minimum edit distance to convert an empty string into the first j characters of string B is j, corresponding to all insertion operations; (3) Edit distance recursion; When i>0 and j>0, if A[i]=B[j], then D[i][j]=D[i-1][j-1], that is, the current character matches and the edit distance is equal to the previous one; if A[i] is not equal to B[j], then D[i][j]=min{D[i-1][j]+1,D[i][j-1]+1,D[i-1][j-1]+1}, which correspond to deletion, insertion and replacement operations respectively, and the minimum value of the three is taken as the current edit distance; (4) Output the edit distance; D[m][n] is the edit distance of converting string A to string B; Traverse the address string A of the investigator, and regard other addresses B whose edit distance to the length of A is less than the threshold (20%) and whose actual distance is less than 300m as similar addresses to address A. Count the proportion of addresses similar to address A to the total number of addresses. If the proportion is greater than the threshold (10%), mark the questionnaire containing address A and its similar addresses as "address similarity pending"; Traverse the investigators and count the proportion of male individual samples in all questionnaires of the investigators. If the male proportion exceeds the threshold (greater than 60% or less than 40%), the investigator is marked as "abnormal".

8. A method for quantitatively evaluating the quality of residents' travel questionnaires according to claim 1, characterized in that: In the S500, the specific steps are: calculating the investigator score according to the investigator's questionnaire marking; setting the investigator's initial score, such as 10 points; For different questionnaire marks, corresponding deduction weights are set. For example, the deduction weights w1, w2, w3, w4, and w5 for the marks of location pending, completeness pending, information rationality pending, travel rationality pending, and address duplication pending are set to -3, -1, -1, -2, and -0.5, respectively. Then, the weighted average method is used to calculate the investigator's deduction value, and the difference between the investigator's initial score and the deduction value is taken as the final score. The calculation formula is: In the formula, totalscore represents the investigator's initial score, N represents the number of questionnaires of the investigator, and x i,j Indicates whether questionnaire i has the jth mark, w j Indicates the weight of the jth mark; the scores are sorted from large to small and displayed as the basis for evaluating investigators.

Images