A method for translating a name of a ship

Abstract

The application discloses a ship name translation method, comprising the following processes: a. receiving input information, dividing the input information into one or more vocabularies, wherein the vocabularies at least include a subject vocabulary; b. comparing each subject vocabulary with data in a preset database in sequence; b1. comparing the subject vocabulary with a geographical name vocabulary in the database, screening a corresponding translation geographical name according to a comparison result, and calculating a first score of each geographical name; b2. judging a location where a ship is newly positioned, and calculating a second score of each geographical name; b3. obtaining a recent trajectory record of the ship, obtaining a plurality of trajectory points on the trajectory record, and calculating a third score of each geographical name; c. calculating a matching score of each geographical name; and selecting a geographical name with the highest matching score as a translation result of the subject vocabulary. The translation method greatly reduces interference information and improves translation accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of ships, in particular to a method for translating ship names. Background Art

[0002] Existing ship names only support the input of letters, numbers and symbols. A ship name is composed of words such as province abbreviations, place name pinyin, and ship types. Since there are partial overlaps in the letter composition of province abbreviation pinyin and place name pinyin, for example, Chongqing Municipality (Yu yu) and Henan Province (Yu yu), Jiangxi Province (Gan gan) and Gansu Province (Gan gan), Hebei Province (Ji ji) and Jilin Province (Ji ji), etc. When translating existing ship names, there are numerous interfering translation items, which causes troubles to ship staff.

[0003] Existing translation software does not specifically translate ship names, and there are the following situations: 1. No translation result, 2. Translation errors of near-homophones or homophones. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for translating ship names, so as to solve the problem of inaccurate translation caused by numerous interferences existing in ship names in the background art.

[0005] In order to achieve the above purpose, the present invention discloses a method for translating ship names, which includes the following process: a. Receiving input information of a certain ship name to be translated, dividing the input information into one or more words, wherein the words include at least one main word.

[0006] b. Comparing each main word with the data in a preset database in sequence, wherein the database stores the mapping relationships among place name words, place name names translated corresponding to the place name words, and the position attributes corresponding to the place name names; the comparison process includes calculating the following first score, and calculating at least one of the second score and the third score.

[0007] b1. Comparing the main word with the place name words in the database, screening out the place name names corresponding to the translations of the place name words according to the comparison result, and calculating the first score of each place name name.

[0008] b2. Determining the position of the latest positioning of the ship, comparing it with the position attribute corresponding to the place name name obtained in process b1, and calculating the second score of each place name name according to the comparison result.

[0009] b3. Obtaining the latest track record of the ship, acquiring several track points on the track record, comparing the position where the track point is located with the position attribute corresponding to the place name name obtained in process b1, and calculating the third score of each place name name according to the comparison result.

[0010] c. After weighting the scores of each place name in process b, obtain the matching score of each place name; select the place name with the highest matching score as the translation result of the main word; if there are multiple main words, then concatenate the place name with the highest matching score of each main word to obtain the translation result.

[0011] Furthermore, it also includes a step to effectively determine the translation results, namely, calculating the total score of the translation results of the main vocabulary through weighted calculation, and judging whether the total score meets the first standard: if the total score exceeds the first standard, the translation result is directly output; if the total score does not exceed the first standard, the translation result is determined to be invalid; if the translation result is determined to be invalid, the translation result is not output, or all place names corresponding to the matching scores are output.

[0012] Preferably, the trajectory point in process b3 is a ship docking point, which is a trajectory point with zero speed in the ship's historical navigation trajectory.

[0013] Preferably, the database also stores the mapping relationship between ship type names and type words. In process a, the input information is divided into main words and type words. The divided type words are compared with the ship type names stored in the database to obtain the ship type name corresponding to the type word as the translation result. The translation results of the main words and the translation results of the type words are concatenated to obtain the translation result of the ship name.

[0014] Furthermore, the input information in process a also includes numerical codes. In process a, the input information is divided into main vocabulary, category vocabulary, and numerical codes. The translation results of the main vocabulary, the translation results of the category vocabulary, and the numerical codes are concatenated to obtain the translation result of the ship name.

[0015] The main vocabulary includes province abbreviations and place name pinyin words; the place names in the database include province abbreviations and region names; the province abbreviations are matched by a weighted calculation of a first score, a second score, and a third score to obtain a matching score, and the place name pinyin words are matched by a weighted calculation of a first score and a second score, or a first score and a third score to obtain a matching score.

[0016] Preferably, the weight of the first score in the abbreviation of the province is set to 0.7 to 0.8, the weight of the second score is set to 0.1, and the remainder is the weight of the third score; the weight of the first score in the pinyin of the place name is set to 0.7 to 0.8, and the remainder is the weight of the second score.

[0017] Preferably, the region names stored in the database include at least three levels: province, city, and district. Region names that do not have rivers or seas, are not adjacent to rivers or seas, or do not have at least one type of ship are not included in the database.

[0018] Preferably, when calculating the first score, the place name pinyin words are compared with place names in the database in descending order of regional administrative level; when calculating the second score, the place names corresponding to the place name pinyin words are scored in descending order of distance from the ship's location; when calculating the third score, the place names corresponding to the place name pinyin words are scored in descending order of distance from the location with the most ship trajectory points.

[0019] In one embodiment, the main vocabulary includes province abbreviations and place name pinyin words; the total score of the translation result of the main vocabulary is F = 0.3S + 0.7D; where S is the matching score of province abbreviations and D is the matching score of place name pinyin words; the first standard is at least 80 points.

[0020] The matching score S of the place names corresponding to the abbreviations of the provinces is calculated as follows: S = score t +score l +score h Among them, score t The first score for the place name corresponding to the abbreviation of the province. l The second score is the score for the place name corresponding to the abbreviation of the province. h The third fraction is the place name corresponding to the abbreviation of the province.

[0021] The first fraction for the place name corresponding to the abbreviation of a province is calculated as follows:

[0022]

[0023] Among them, C1C2…C n The abbreviation for province, M P This refers to the place name of province P corresponding to the abbreviation of a province in the database;

[0024] The calculation method for the second fraction of the place name corresponding to the abbreviation of the province is as follows:

[0025]

[0026] The third fraction for place names corresponding to abbreviated provincial terms is calculated as follows:

[0027]

[0028] Where A represents the set of trajectory points of the ship, n represents the number of points of trajectory points within province P, and m represents the total number of points in A.

[0029] The matching score D = score of the place name pinyin words to the corresponding place name names trans +score loc ;

[0030] Among them, score trans The score is the first score for the place name corresponding to the pinyin words of the place name. loc The third fraction is the place name corresponding to the place name in the pinyin words.

[0031] The method for calculating the first fraction of the place name corresponding to the pinyin words is as follows:

[0032]

[0033] Where q represents the number of place names matched in the database using the pinyin words for place names; the location attribute of each place name is a place name located within province P.

[0034] The calculation method for the third score of the place name corresponding to the place name in pinyin is as follows: if q = 1, the third score is 20 points. If q > 1, the location attribute corresponding to the place name is matched with the location with the most ship docking points in the ship trajectory points. The location with the most ship docking points is sorted from near to far with the location attribute corresponding to the place name, with a maximum of 6 positions. The third score decreases from 20 points.

[0035] Compared with existing technologies, the present invention has the following advantages:

[0036] By establishing a database that stores place name terms, corresponding place name translations, and location attributes of the corresponding place names, the vocabulary of ship names is compared with the database. Based on the degree of matching with the place name, and by adding the ship's latest location and trajectory records, the place name with the highest matching score is selected as the translation result of the ship name. This greatly reduces interference information, improves translation accuracy, and provides convenience for people working in the shipping industry. Detailed Implementation

[0037] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to specific embodiments.

[0038] This invention discloses a method for translating ship names. Ship names generally consist of abbreviations of the province, pinyin of the place name, and a term indicating the type of ship (e.g., MINLONGYU), or abbreviations of the province, pinyin of the place name, a term indicating the type of ship, and a numerical designation (e.g., MINLONGYU223). The abbreviations of the province and pinyin of the place name are the primary objects of translation in this invention, i.e., the main vocabulary.

[0039] The translation method of the present invention includes the following process:

[0040] a. Vocabulary segmentation

[0041] The system receives input information about a ship name to be translated and divides the input information into one or more terms, each term including at least one main term. In this embodiment, the main terms include two: the first is a province abbreviation term, and the second is a place name pinyin term, which is located between the province abbreviation term and the category term.

[0042] The database stores the mapping relationship between place name vocabulary, the corresponding place name translations, and the location attributes of the corresponding place names. The place names correspond to the abbreviations of provinces and the names of regions. It also stores the mapping relationship between ship type names and type vocabulary.

[0043] For example:

[0044] The input information is MINLONGYU, which can be divided into two main words: MIN and LONG. MIN is the abbreviation of the province, LONG is the pinyin of the place name, and YU is the category word.

[0045] The input information is YUYUNYU80418. This input information can be divided into two main words: YU and YUN. YU is the abbreviation of the province, YUN is the pinyin of the place name, YU is the category word, and 80417 is the number code.

[0046] The input information is YUEZHAOQINGHUO261. This input information can be divided into two main words: YUE and ZHAOQING. YUE is the abbreviation of the province, ZHAOQING is the pinyin of the place name, HUO is the category word, and 80417 is the number code.

[0047] The vocabulary for ship types is relatively fixed. The existing mapping relationship between ship type vocabulary and ship type names is as follows: cargo (HUO), fishing (YU), aquaculture (YUYANG), fish transport (YUYUN), fishing command (YUZHI), fish oil (YUYOU), fish water (YUSHUI), fish cold storage (YULENG). Because there is a one-to-one correspondence between the translation results of ship type vocabulary and the input information, when the spelling of one ship type vocabulary is input, only one translation result is output.

[0048] To reduce distracting options, place names that do not have a river or sea area, are not adjacent to a river or sea area, or do not have at least one type of ship are not included in the database, such as Gan, Ji, and Yu, in order to exclude distracting translation options.

[0049] b. Vocabulary comparison

[0050] Each subject term is sequentially compared with data in a pre-defined database. The comparison process includes calculating a first score and at least one of a second or third score. Calculating the first score for the subject terms is a necessary step in this application's technical solution, while calculating the second and third scores is an optional step; either a single second or third score can be calculated, or both can be calculated.

[0051] b1. Compare the main vocabulary with the place name vocabulary in the database, select the place names corresponding to the place name vocabulary based on the comparison results, and calculate the first score of each place name.

[0052] b2. Determine the ship's latest location and compare it with the location attributes corresponding to the place names obtained in process b1. Based on the comparison results, calculate the second score for each place name. Here, the latest location can be determined by drawing a circle within a certain range (e.g., a circle with a radius of 100 nautical miles) based on the ship's latest location point, and then matching the place names of cities and districts within that range.

[0053] b3. Obtain the ship's most recent trajectory record. Extract several trajectory points from this record. Compare the locations of these trajectory points with the location attributes corresponding to the place names obtained in process b1. Based on the comparison results, calculate the third score for each place name. Preferably, the trajectory points can be ship berthing points, which are the trajectory points in the ship's historical navigation trajectory where the speed is zero.

[0054] For example, when abbreviated terms correspond to abbreviated province names, a first score, a second score, and a third score can be calculated. Similarly, when a place name's pinyin corresponds to a region name, a first score and a second score, or a first score and a third score, can be calculated. When calculating the first score, the place name's pinyin is compared to place names in the database in descending order of administrative level. When calculating the second score, the place name corresponding to the pinyin is scored in descending order of distance from the ship's location. When calculating the third score, the place name is scored in descending order of distance from the location with the most ship tracking points. By comparing and scoring according to this order or rule, faster and more accurate translations can be achieved.

[0055] c. Calculate the matching score

[0056] The scores of each place name in process b are weighted to obtain the matching score of each place name; the place name with the highest matching score is selected as the translation result of the main word; if there are multiple main words, the place name with the highest matching score of each main word is concatenated to obtain the translation result.

[0057] The weighted calculation method involves designing weights for each score based on practical experience, and then adding them together to obtain the matching score. For example, in the case of province abbreviations, the weight of the first score is set to 0.7-0.8, the weight of the second score is set to 0.1, and the remainder is the weight of the third score; in the case of place name pinyin words, the weight of the first score is set to 0.7-0.8, and the remainder is the weight of the second score.

[0058] d. Validity determination and output of translation results.

[0059] The total score of the translation results of the main vocabulary is calculated by weighted operation. The total score is judged to see if it meets the first standard: if the total score exceeds the first standard, the translation result is output directly; if the total score does not exceed the first standard, the translation result is determined to be invalid; if the translation result is determined to be invalid, no translation result is output, or all place names corresponding to the matching scores are output.

[0060] This step can be configured as needed. Its purpose is that when the total score of the translation is too low, the system determines that the translation accuracy is poor. At this time, it can be declared invalid or the place names corresponding to all matching scores can be output for subsequent manual evaluation.

[0061] When the input information includes a main vocabulary and a type vocabulary, the divided type vocabulary is compared with the ship type names stored in the database, the corresponding ship type name of the type vocabulary is obtained as the translation result, and the translation result of the main vocabulary and the translation result of the type vocabulary are spliced to obtain the translation result of the ship name. For example: if the input information is MINLONGYU, the output translation result is: 闽龙渔.

[0062] When the input information includes a main vocabulary, a type vocabulary and a number, the translation result of the main vocabulary, the translation result of the type vocabulary and the number are spliced to obtain the translation result of the ship name. For example: if the input information is YUYUNYU80418, the output translation result is: 闽龙渔80418; if the input information is YUYUNYU80418, the output translation result is: 渝云渔80418; if the input information is YUEZHAOQINGHUO261, the output translation result is: 粤肇庆货0261.

[0063] Embodiment 1

[0064] This embodiment is a specific translation method for provincial abbreviation vocabularies, and this embodiment takes the provincial abbreviation vocabulary "YU" in the input information YUYUNYU80418 as an example for description.

[0065] The calculation process of three types of scores corresponding to the provincial abbreviation name of the provincial abbreviation vocabulary is described in detail as follows.

[0066] (1) The calculation method of the first score of the provincial abbreviation vocabulary is: compare the provincial abbreviation vocabulary with the database, and when the letter composition of the provincial abbreviation vocabulary completely matches that in the database, record the first score of the corresponding provincial abbreviation name as score t of 70 points; if there is one unmatched character, reduce the score by 10 points, which is shown in the following formula.

[0067]

[0068] Wherein, C1C2...C n represents the provincial abbreviation vocabulary, P is the provincial abbreviation name of the latest positioning position of the ship, M is the mapping relationship between the provincial abbreviation name and the provincial abbreviation vocabulary, M P represents the place name of the province P corresponding to the provincial abbreviation vocabulary in the database; wherein the mapping relationship of M is M = {("Chongqing City", "渝", "YU"), ("Jiangxi Province", "赣", "GAN"), ("Hebei Province", "冀", "JI"), ("Henan Province", "豫", "YU")...}

[0069] (2) The calculation method of the second score of the place name corresponding to the provincial abbreviation vocabulary is:

[0070]

[0071] (3) The calculation method of the third score corresponding to the place name of the province abbreviation word is:

[0072]

[0073] Wherein, A represents the set of trajectory points of the ship, n represents the number of trajectory points located within province P, and m represents the total number of points in A;

[0074] For example, in this embodiment, the province abbreviation word is "YU". When calculating the first score, there are multiple translated place names. For example, YU can be translated as Yu (Chongqing Municipality) or Yu (Henan Province), so the first score of both "Yu" and "Yu" is 70 points. According to the latest positioning of the ship, if the latest positioning of the ship is in Chongqing Municipality, the second score of "Yu" (Chongqing) is 10 points, and the second score of "Yu" (Henan) is 5 points. Further combined with the ship's docking points, assuming the preset time is 7 days, there are 10 ship docking points within 7 days, 8 of which are located in Chongqing Municipality and 2 are located in Hubei Province, then the third score of "Yu" (Chongqing) is 16 points, while the third score of "Yu" (Henan) is 0 points.

[0075] Therefore: the matching score S of the province abbreviation name "Yu" (Chongqing) = score t +score l +score h =70+10+16=96. The matching score S of the province abbreviation name "Yu" (Henan) = score t +score l +score h =70+5+0=75.

[0076] The "Yu" (Chongqing) with the highest matching score is selected as the translation result of the main word "YU".

[0077] Second Embodiment

[0078] This embodiment is a specific translation method for pinyin abbreviation words of place names. This embodiment takes the place name abbreviation word "YUN" in the input information YUYUNYU80418 as an example for description.

[0079] The matching score D of the place name pinyin word corresponding to the place name = score trans +score loc ;

[0080] Wherein, score trans is the first score of the place name pinyin word corresponding to the place name, and score loc is the third score of the place name pinyin word corresponding to the place name.

[0081] (1) The calculation method of the first score of the place name pinyin vocabulary corresponding to the place name name is:

[0082]

[0083] where q is the number of place name names matched by the place name pinyin vocabulary in the database. The place name names in the database at this time only include the place name names in the province where the first main vocabulary is located. Since the first main vocabulary "YU" is matched to "Yu" (representing Chongqing) in this embodiment, the location attributes of the matched place name names are all place name names located in Chongqing.

[0084] (2) The calculation method of the third score of the place name pinyin vocabulary corresponding to the place name name is: if q=1, the third score is 20 points; if q>1, the location attribute corresponding to the place name name is matched with the location with the largest number of ship stop points in the ship trajectory points, the locations with the largest number of ship stop points and the location attributes corresponding to the place name names are sorted from the nearest to the farthest, and the third score decreases from 20 points downward, such as 18 points, 16 points, 14 points, 12 points, 10 points

[0085] For example, in this embodiment, the only place name name matched by "YUN" is "Yun" (Yunyang District, Yunyang County), that is, q=1, so the matching score D of the place name pinyin abbreviation vocabulary corresponding to the place name name "Yun" is score trans +score loc =80+20=100.

[0086] The "Yun" with the highest matching score is selected as the translation result of the main vocabulary "YUN".

[0087] Third Embodiment

[0088] This embodiment discloses a specific method for judging the validity of the translation result. As shown in the first embodiment and the second embodiment, after translating the two main vocabularies, it is necessary to judge their validity. The total score of the translation result of the main vocabulary is calculated through weighted operation, and it is judged whether the total score reaches the first standard.

[0089] For example, in this embodiment, the total score F=0.3S+0.7D; where S is the matching score of the province abbreviation vocabulary, D is the matching score of the place name pinyin vocabulary; the first standard is 80 points.

[0090] According to the scores of the first embodiment and the second embodiment, S=75, D=100, then the total score F=0.3×75+0.7×100=92.5>80, it can be seen that the total score is greater than the first standard, so the translation result is a valid result and is output directly.

[0091] If the total score is less than 80 points, it is determined that the translation result is invalid, then all translation results corresponding to the matching scores are output, and processed by manual judgment at the client.

[0092] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for translating ship names, characterized in that, Includes the following processes: a. Receive input information of a ship name to be translated, and divide the input information into one or more words, wherein the words include at least one main word; b. Compare each subject term sequentially with data in a pre-set database, which stores mappings of place name terms, their translated place names, and the location attributes of the corresponding place names. The comparison process includes calculating a first score and at least one of a second or third score: b1. Compare the main vocabulary with the place name vocabulary in the database, and select the place names that correspond to the place name vocabulary based on the comparison results. Calculate the first score of each place name. b2. Determine the latest location of the ship and compare it with the location attribute corresponding to the place name obtained in process b1. Calculate the second score of each place name based on the comparison result. b3. Obtain the ship's most recent trajectory record, acquire several trajectory points from the trajectory record, compare the location of the trajectory points with the location attributes corresponding to the place names obtained in process b1, and calculate the third score of each place name based on the comparison results. The main vocabulary includes abbreviations of provinces and pinyin words for place names; the place names in the database include abbreviations of provinces and regional names. The matching score for the abbreviation of a province and the abbreviation of a province is obtained by weighting the first score, the second score, and the third score. The matching score for the pinyin of a place name and the place name is obtained by weighting the first score and the second score, or the first score and the third score. The weight of the first score in the abbreviation of the province is set to 0.7-0.8, the weight of the second score is set to 0.1, and the remainder is the weight of the third score; the weight of the first score in the pinyin of the place name is set to 0.7-0.

8. c. After weighting the scores of each place name in process b, obtain the matching score of each place name; select the place name with the highest matching score as the translation result of the main word; if there are multiple main words, then concatenate the place name with the highest matching score of each main word to obtain the translation result.

2. The method for translating ship names according to claim 1, characterized in that: It also includes a step to effectively judge the translation results, namely, to calculate the total score of the translation results of the main vocabulary through weighted calculation, and to judge whether the total score meets the first standard: if the total score exceeds the first standard, the translation result is directly output; If the total score does not exceed the first standard, the translation result is deemed invalid; if the translation result is deemed invalid, no translation result is output, or all place names corresponding to the matching scores are output.

3. The method for translating ship names according to claim 1, characterized in that: The trajectory point mentioned in process b3 is the ship's docking point, which is the trajectory point with zero speed in the ship's historical navigation trajectory.

4. The method for translating ship names according to claim 1, characterized in that: The database also stores the mapping relationship between ship type names and type vocabulary. In process a, the input information is divided into main vocabulary and type vocabulary. The divided type vocabulary is compared with the ship type names stored in the database to obtain the ship type name corresponding to the type vocabulary as the translation result. The translation results of the main vocabulary and the type vocabulary are concatenated to obtain the translation result of the ship name.

5. The method for translating ship names according to claim 4, characterized in that: The input information in process a also includes numerical codes. In process a, the input information is divided into main vocabulary, category vocabulary and numerical codes. The translation results of the main vocabulary, the translation results of the category vocabulary and the numerical codes are concatenated to obtain the translation result of the ship name.

6. The method for translating ship names according to claim 1, characterized in that: The database stores at least three levels of regional names: province, city, and district. Regional names that do not have rivers or seas, are not adjacent to rivers or seas, or do not have at least one type of ship are not included in the database.

7. The method for translating ship names according to claim 1, characterized in that: When calculating the first score, the place name pinyin words are compared with the place name names in the database in descending order of the regional administrative level of the place name; When calculating the second score, the place names corresponding to the place name pinyin words are scored in descending order of distance from the ship's location. When calculating the third score, the place name corresponding to the place name in pinyin is scored in descending order of distance from the location with the most ship track points.

8. The method for translating ship names according to claim 2, characterized in that: The total score for the translation of the main vocabulary is F = 0.3S + 0.7D; where S is the score for matching the abbreviation of the province and D is the score for matching the pinyin of the place name; the first standard is at least 80 points. The matching score S of the place names corresponding to the abbreviations of the provinces is calculated as follows: ; in, The first fraction is the place name corresponding to the abbreviation of the province. The second fraction is the place name corresponding to the abbreviation of the province. The third fraction is the place name corresponding to the abbreviation of the province; The first fraction for the place name corresponding to the abbreviation of a province is calculated as follows: in, Abbreviations for provinces This refers to the place name of province P corresponding to the abbreviation of a province in the database; The calculation method for the second fraction of the place name corresponding to the abbreviation of the province is as follows: ； The third fraction for place names corresponding to abbreviated provincial terms is calculated as follows: ； Where A represents the set of trajectory points of the ship, n represents the number of trajectory points within province P, and m represents the total number of points in A; The matching score of the place name pinyin words to the corresponding place names ; in, The first fraction is the place name corresponding to the place name in pinyin. The third fraction is the place name corresponding to the place name in pinyin words; The method for calculating the first fraction of the place name corresponding to the pinyin words is as follows: ； Where q represents the number of place names matched in the database using the pinyin words for place names; the location attribute of each place name is a place name located within province P. The calculation method for the third score of the place name corresponding to the pinyin words is as follows: if q=1, the third score is 20 points. If q>1, the location attribute corresponding to the place name is matched with the location with the most ship docking points in the ship trajectory points. The location with the most ship docking points is sorted from near to far with the location attribute corresponding to the place name, with a maximum of 6 positions. The third score decreases from 20 points.

Images