CN105335456A - Relevancy priority ordering method used for environmental protection regulation retrieval - Google Patents

Abstract

Translated fromChinese

本发明涉及一种用于环境保护法规检索的关联优先排序方法，属于知识发现领域。本发明首先对环保法律法规检索系统构建一个关键词表、关键字表；然后清洗用户输入的数据并提炼候选词；最后根据候选词的个数，计算距离并排序输出。本发明采用索引的方式，将庞大的信息源提炼成一个关键词表，作为整个信息源的目录索引，只要与索引匹配查询便能快速的在庞大的信息源中找到有意义的信息，进一步提高检索效率；采用计算整个关键词库中各个独立汉字之间的距离，将其距离值存储在关键字表中，因此在查询匹配的时候就只需要去寻找距离值最小的元素就能找到关联度很高的词语或词组；在提高检索效率的同时，也提高了检索结果与搜索意图之间关联度的准确性。The invention relates to a correlation prioritization method for retrieval of environmental protection laws and regulations, belonging to the field of knowledge discovery. The invention first constructs a keyword table and a keyword table for the retrieval system of environmental protection laws and regulations; then cleans the data input by the user and extracts candidate words; finally calculates the distance and sorts the output according to the number of candidate words. The present invention adopts the method of indexing to extract huge information sources into a keyword table, which is used as a directory index of the entire information source. As long as it is matched with the index, it can quickly find meaningful information in the huge information sources, further improving Retrieval efficiency: calculate the distance between independent Chinese characters in the entire keyword database, and store the distance value in the keyword table, so when querying and matching, you only need to find the element with the smallest distance value to find the degree of association Very high words or phrases; while improving retrieval efficiency, it also improves the accuracy of the correlation between retrieval results and search intent.

Description

Translated fromChinese

一种用于环境保护法规检索的关联优先排序方法A Correlation Prioritization Method for Retrieval of Environmental Protection Regulations

技术领域technical field

本发明涉及一种用于环境保护法规检索的关联优先排序方法，属于知识发现领域。The invention relates to a correlation prioritization method for retrieval of environmental protection laws and regulations, belonging to the field of knowledge discovery.

背景技术Background technique

信息爆炸是当今信息社会的一大特点，从web上进行搜索会查询到大量冗余繁琐信息，需要我们再逐一去筛选来获得我们想要的信息。因而如何快速找到一种方法，给用户更简洁的呈现出更有意义的信息成为了一个关键的问题。因此，为解决这一问题，提出知识发现，知识发现是从数据集中识别出有效的、新颖的、潜在有用的，以及最终可理解的模式的非平凡过程。目的是向使用者屏蔽原始数据的繁琐细节，从原始数据中提炼出有意义的、简洁的知识，直接向使用者报告。为了向使用者提供更有意义的信息，本方法被提出来，它通过计算元素与元素之间的距离，即关联度，以最快的方式寻找到与使用者想搜索的信息的距离最优的词语组合，然后对应索引目录快速准确查找出更有意义的信息，即用户所需要信息。Information explosion is a major feature of today's information society. Searching from the web will result in a large amount of redundant and cumbersome information, which requires us to filter one by one to obtain the information we want. Therefore, how to quickly find a way to present more concise and meaningful information to users has become a key issue. Therefore, to address this issue, knowledge discovery is proposed, which is the non-trivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns from data sets. The purpose is to shield the cumbersome details of the original data from the user, extract meaningful and concise knowledge from the original data, and report directly to the user. In order to provide users with more meaningful information, this method is proposed. It calculates the distance between elements, that is, the degree of association, and finds the optimal distance to the information that users want to search for in the fastest way. Word combinations, and then quickly and accurately find more meaningful information corresponding to the index directory, that is, the information users need.

发明内容Contents of the invention

本发明提供了一种用于环境保护法规检索的关联优先排序方法，以用于解决快速查找用户所需要信息的问题。The invention provides an association prioritization method for retrieval of environmental protection laws and regulations to solve the problem of quickly finding information required by users.

本发明的技术方案是：一种用于环境保护法规检索的关联优先排序方法，首先对环保法律法规检索系统构建一个关键词表A、关键字表B；然后清洗用户输入的数据并提炼候选词；最后根据候选词的个数，计算距离并排序输出。The technical solution of the present invention is: a method for prioritizing associations for retrieval of environmental protection laws and regulations, first constructing a keyword table A and keyword table B for the retrieval system of environmental protection laws and regulations; then cleaning the data input by the user and refining candidate words ; Finally, according to the number of candidate words, calculate the distance and sort the output.

所述用于环境保护法规检索的关联优先排序方法的具体步骤如下：The specific steps of the associated prioritization method for retrieval of environmental protection regulations are as follows:

Step1、首先建立系统模型：Step1, first establish a system model:

对环保法律法规检索系统构建一个关键词表A、关键字表B；其中，关键词表A：存储着法规名称及法规中抽取出来的t组关键词；关键字表B：存储着关键词表A中每个关键词拆分成的不同字m个及各个字之间的特征值Aij；Aij表示角标为i和j所代表的字的组合出现在关键词表A中的频数，角标i、j为关键词表A中每个关键词拆分成的不同字在关键字表B中的标记；Construct a keyword table A and keyword table B for the retrieval system of environmental protection laws and regulations; among them, keyword table A: store the name of the law and t groups of keywords extracted from the law; keyword table B: store the keyword table Each keyword in A is divided into m different words and the characteristic value Aij between each word; Aij indicates the frequency of the combination of words represented by subscripts i and j appearing in keyword table A, and subscripts i, j are the marks in the keyword table B of the different words that each keyword is split into in the keyword table A;

Step2、清洗用户输入的数据并提炼候选词：Step2. Clean the data entered by the user and refine the candidate words:

针对用户输入的数据进行分词并去除停用词，将剩余的分词作为候选词；Segment the data entered by the user and remove stop words, and use the remaining word segmentation as candidate words;

Step3、根据候选词的个数，计算距离并排序输出：Step3. According to the number of candidate words, calculate the distance and sort the output:

Step3.1、若候选词个数为1时：Step3.1, if the number of candidate words is 1:

从关键字表B中获取与候选词的首字x联结的字、首字x之间的特征值Aix、获取尾字y、与尾字y联结的字之间的特征值Ayj；计算Aix≠0情况下首字与关键字表B中字的距离dix且得到ixy对应的词组合，计算Ayj≠0情况下尾字与关键字表B中字的距离dyj且得到xyj对应的词组合；根据dix、dyj从小到大的顺序排列其对应的词组合；根据词组合的顺序，将词组合与关键词表A中的关键词进行匹配获取对应的法规名称，将匹配的结果去除重复后按照顺序显示；其中，当出现dix＝dyj，则dix、dyj对应的词组合进行随机排序；Obtain from keyword table B the word that is connected with the first word x of the candidate word, the characteristic value Aix between the first word x, obtain the last word y, and the characteristic value Ayj between the words connected with the last word y; calculate Aix≠ In the case of 0, the distance dix between the first character and the word in the keyword table B is obtained and the word combination corresponding to ixy is obtained, and the distance dyj between the last word and the word in the keyword table B is calculated under the situation of Ayj ≠ 0 and the word combination corresponding to xyj is obtained; according to dix, dyj arranges the corresponding word combinations in order from small to large; according to the order of the word combinations, match the word combinations with the keywords in the keyword table A to obtain the corresponding legal names, remove the duplicates of the matching results and display them in order; Wherein, when dix=dyj occurs, then the word combinations corresponding to dix and dyj are randomly sorted;

Step3.2、若候选词个数不为1时：Step3.2, if the number of candidate words is not 1:

将多个候选词按输入顺序排列，分别计算相邻两个候选词中先输入的候选词的尾字u与后输入的候选词的首字v的距离duv及对应的两个候选词构成的词组合；从关键字表B中获取与各个候选词的首字x联结的字、首字x之间的特征值Aix、获取尾字y、与尾字y联结的字之间的特征值Ayj；计算Aix≠0情况下首字与关键字表B中字的距离dix且得到ixy对应的词组合，计算Ayj≠0情况下尾字与关键字表B中字的距离dyj且得到xyj对应的词组合；根据duv、dix、dyj从小到大的顺序排列其对应的词组合；根据词组合的顺序，将词组合与关键词表A中的关键词进行匹配获取对应的法规名称，将匹配的结果去除重复后按照顺序显示；其中，当出现duv＝dix＝dyj，则仅仅保留duv对应的词组合进行排序，当出现dix＝dyj，则dix、dyj对应的词组合进行随机排序；Arrange a plurality of candidate words in the input order, respectively calculate the distance duv between the last word u of the candidate word input first and the first word v of the candidate word input later among the two adjacent candidate words and the corresponding two candidate words Word combination; from the keyword table B, the character value Aix between the first word x and the first word x of each candidate word is obtained, and the end word y is obtained, and the characteristic value Ayj between the words connected with the last word y ; Calculate the distance dix between the first word and the word in the keyword table B under the condition of Aix≠0 and obtain the word combination corresponding to ixy, calculate the distance dyj between the last word and the word in the keyword table B under the situation of Ayj≠0 and obtain the word combination corresponding to xyj ; According to the order of duv, dix, dyj from small to large, arrange their corresponding word combinations; according to the order of word combinations, match the word combinations with the keywords in keyword table A to obtain the corresponding legal name, and remove the matching result Display in order after repetition; wherein, when duv=dix=dyj occurs, only the word combinations corresponding to duv are reserved for sorting, and when dix=dyj occurs, then the word combinations corresponding to dix and dyj are randomly sorted;

所述u、v、x、y为字在关键字表B中的标记。The u, v, x, and y are marks of words in the keyword table B.

所述$duv=\frac{1+e^{-Auv}}{t},dix=\frac{1+e^{-Aix}}{t},dyj=\frac{1+e^{-Ayj}}{t};$其中Auv、Aix、Ayj分别表示角标为u、v所代表的字的组合，角标为i、x所代表的字的组合，角标为y、j所代表的字的组合出现在关键词表A中的频数；duv、dix、dyj分别表示角标为u、v所代表的字，角标为i、x所代表的字，角标为y、j所代表的字的距离。said$duv=\frac{1+e^{-Auv}}{t},dix=\frac{1+e^{-Aix}}{t},d\mathrm{the\; y}j=\frac{1+e^{-A\mathrm{the\; y}j}}{t};$ Among them, Auv, Aix, and Ayj represent the combination of characters represented by u and v respectively, the combination of characters represented by i and x, and the combination of characters represented by y and j appear in keywords The frequency numbers in Table A; duv, dix, and dyj represent the distance between the characters represented by u and v, the words represented by i and x, and the words represented by y and j respectively.

本发明的有益效果是：The beneficial effects of the present invention are:

采用索引的方式，将庞大的信息源提炼成一个关键词表，作为整个信息源的目录索引。因此，只要与索引匹配查询便能快速的在庞大的信息源中找到有意义的信息，进一步提高检索效率。Using the index method, the huge information source is refined into a keyword list, which is used as the catalog index of the entire information source. Therefore, as long as the query is matched with the index, meaningful information can be quickly found in the huge information source, further improving the retrieval efficiency.

采用计算整个关键词库中各个独立汉字之间的距离，将其距离值存储在关键字表中。因此在查询匹配的时候就只需要去寻找距离值最小的元素就能找到关联度很高的词语或词组。在提高检索效率的同时，也提高了检索结果与搜索意图之间关联度的准确性。Calculate the distance between each independent Chinese character in the entire keyword database, and store the distance value in the keyword table. Therefore, when querying and matching, you only need to find the element with the smallest distance value to find words or phrases with high correlation. While improving the retrieval efficiency, the accuracy of the correlation degree between the retrieval result and the search intent is also improved.

附图说明Description of drawings

图1为本发明元素间距离网状示意图；Fig. 1 is the network schematic diagram of the distance between elements of the present invention;

图2为本发明元素间距离网状实例示意图。Fig. 2 is a schematic diagram of an example of a mesh-like distance between elements in the present invention.

具体实施方式detailed description

实施例1：如图1-2所示，一种用于环境保护法规检索的关联优先排序方法，首先对环保法律法规检索系统构建一个关键词表A、关键字表B；然后清洗用户输入的数据并提炼候选词；最后根据候选词的个数，计算距离并排序输出。Embodiment 1: As shown in Fig. 1-2, a kind of association prioritization method that is used for the retrieval of environmental protection laws and regulations, first builds a keyword table A, keyword table B to the environmental protection laws and regulations retrieval system; Then cleans the user input Data and extract candidate words; finally, according to the number of candidate words, calculate the distance and sort the output.

所述用于环境保护法规检索的关联优先排序方法的具体步骤如下：The specific steps of the associated prioritization method for retrieval of environmental protection regulations are as follows:

Step1、首先建立系统模型：Step1, first establish a system model:

对环保法律法规检索系统构建一个关键词表A、关键字表B；其中，关键词表A：存储着法规名称及法规中抽取出来的t组关键词；关键字表B：存储着关键词表A中每个关键词拆分成的不同字m个及各个字之间的特征值Aij；Aij表示角标为i和j所代表的字的组合出现在关键词表A中的频数，角标i、j为关键词表A中每个关键词拆分成的不同字在关键字表B中的标记；Construct a keyword table A and keyword table B for the retrieval system of environmental protection laws and regulations; among them, keyword table A: store the name of the law and t groups of keywords extracted from the law; keyword table B: store the keyword table Each keyword in A is divided into m different words and the characteristic value Aij between each word; Aij indicates the frequency of the combination of words represented by subscripts i and j appearing in keyword table A, and subscripts i, j are the marks in the keyword table B of the different words that each keyword is split into in the keyword table A;

Step2、清洗用户输入的数据并提炼候选词：Step2. Clean the data entered by the user and refine the candidate words:

针对用户输入的数据进行分词并去除停用词，将剩余的分词作为候选词；Segment the data entered by the user and remove stop words, and use the remaining word segmentation as candidate words;

Step3、根据候选词的个数，计算距离并排序输出：Step3. According to the number of candidate words, calculate the distance and sort the output:

Step3.1、若候选词个数为1时：Step3.1, if the number of candidate words is 1:

从关键字表B中获取与候选词的首字x联结的字、首字x之间的特征值Aix、获取尾字y、与尾字y联结的字之间的特征值Ayj；计算Aix≠0情况下首字与关键字表B中字的距离dix且得到ixy对应的词组合，计算Ayj≠0情况下尾字与关键字表B中字的距离dyj且得到xyj对应的词组合；根据dix、dyj从小到大的顺序排列其对应的词组合；根据词组合的顺序，将词组合与关键词表A中的关键词进行匹配获取对应的法规名称，将匹配的结果去除重复后按照顺序显示；其中，当出现dix＝dyj，则dix、dyj对应的词组合进行随机排序；Obtain from keyword table B the word that is connected with the first word x of the candidate word, the characteristic value Aix between the first word x, obtain the last word y, and the characteristic value Ayj between the words connected with the last word y; calculate Aix≠ In the case of 0, the distance dix between the first character and the word in the keyword table B is obtained and the word combination corresponding to ixy is obtained, and the distance dyj between the last word and the word in the keyword table B is calculated under the situation of Ayj ≠ 0 and the word combination corresponding to xyj is obtained; according to dix, dyj arranges the corresponding word combinations in order from small to large; according to the order of the word combinations, match the word combinations with the keywords in the keyword table A to obtain the corresponding legal names, remove the duplicates of the matching results and display them in order; Wherein, when dix=dyj occurs, then the word combinations corresponding to dix and dyj are randomly sorted;

Step3.2、若候选词个数不为1时：Step3.2, if the number of candidate words is not 1:

将多个候选词按输入顺序排列，分别计算相邻两个候选词中先输入的候选词的尾字u与后输入的候选词的首字v的距离duv及对应的两个候选词构成的词组合；从关键字表B中获取与各个候选词的首字x联结的字、首字x之间的特征值Aix、获取尾字y、与尾字y联结的字之间的特征值Ayj；计算Aix≠0情况下首字与关键字表B中字的距离dix且得到ixy对应的词组合，计算Ayj≠0情况下尾字与关键字表B中字的距离dyj且得到xyj对应的词组合；根据duv、dix、dyj从小到大的顺序排列其对应的词组合；根据词组合的顺序，将词组合与关键词表A中的关键词进行匹配获取对应的法规名称，将匹配的结果去除重复后按照顺序显示；其中，当出现duv＝dix＝dyj，则仅仅保留duv对应的词组合进行排序，当出现dix＝dyj，则dix、dyj对应的词组合进行随机排序；Arrange a plurality of candidate words in the input order, respectively calculate the distance duv between the last word u of the candidate word input first and the first word v of the candidate word input later among the two adjacent candidate words and the corresponding two candidate words Word combination; from the keyword table B, the character value Aix between the first word x and the first word x of each candidate word is obtained, and the end word y is obtained, and the characteristic value Ayj between the words connected with the last word y ; Calculate the distance dix between the first word and the word in the keyword table B under the condition of Aix≠0 and obtain the word combination corresponding to ixy, calculate the distance dyj between the last word and the word in the keyword table B under the situation of Ayj≠0 and obtain the word combination corresponding to xyj ; According to the order of duv, dix, dyj from small to large, arrange their corresponding word combinations; according to the order of word combinations, match the word combinations with the keywords in keyword table A to obtain the corresponding legal name, and remove the matching result Display in order after repetition; wherein, when duv=dix=dyj occurs, only the word combinations corresponding to duv are reserved for sorting, and when dix=dyj occurs, then the word combinations corresponding to dix and dyj are randomly sorted;

所述u、v、x、y为字在关键字表B中的标记。The u, v, x, and y are marks of words in the keyword table B.

所述$duv=\frac{1+e^{-Auv}}{t},dix=\frac{1+e^{-Aix}}{t},dyj=\frac{1+e^{-Ayj}}{t};$其中Auv、Aix、Ayj分别表示角标为u、v所代表的字的组合，角标为i、x所代表的字的组合，角标为y、j所代表的字的组合出现在关键词表A中的频数；duv、dix、dyj分别表示角标为u、v所代表的字，角标为i、x所代表的字，角标为y、j所代表的字的距离。said$duv=\frac{1+e^{-Auv}}{t},dix=\frac{1+e^{-Aix}}{t},d\mathrm{the\; y}j=\frac{1+e^{-A\mathrm{the\; y}j}}{t};$ Among them, Auv, Aix, and Ayj represent the combination of characters represented by u and v respectively, the combination of characters represented by i and x, and the combination of characters represented by y and j appear in keywords The frequency numbers in Table A; duv, dix, and dyj represent the distance between the characters represented by u and v, the words represented by i and x, and the words represented by y and j respectively.

实施例2：如图1-2所示，一种用于环境保护法规检索的关联优先排序方法，首先对环保法律法规检索系统构建一个关键词表A、关键字表B；然后清洗用户输入的数据并提炼候选词；最后根据候选词的个数，计算距离并排序输出。Embodiment 2: As shown in Fig. 1-2, a kind of association prioritization method that is used for the retrieval of environmental protection laws and regulations, first constructs a keyword table A, keyword table B to the environmental protection laws and regulations retrieval system; Then cleans the user input Data and extract candidate words; finally, according to the number of candidate words, calculate the distance and sort the output.

所述用于环境保护法规检索的关联优先排序方法的具体步骤如下：The specific steps of the associated prioritization method for retrieval of environmental protection regulations are as follows:

Step1、首先建立系统模型：Step1, first establish a system model:

对环保法律法规检索系统构建一个关键词表A、关键字表B；其中，关键词表A：存储着法规名称及法规中抽取出来的t组关键词；关键字表B：存储着关键词表A中每个关键词拆分成的不同字m个及各个字之间的特征值Aij；Aij表示角标为i和j所代表的字的组合出现在关键词表A中的频数，角标i、j为关键词表A中每个关键词拆分成的不同字在关键字表B中的标记；Construct a keyword table A and keyword table B for the retrieval system of environmental protection laws and regulations; among them, keyword table A: store the name of the law and t groups of keywords extracted from the law; keyword table B: store the keyword table Each keyword in A is divided into m different words and the characteristic value Aij between each word; Aij indicates the frequency of the combination of words represented by subscripts i and j appearing in keyword table A, and subscripts i, j are the marks in the keyword table B of the different words that each keyword is split into in the keyword table A;

Step2、清洗用户输入的数据并提炼候选词：Step2. Clean the data entered by the user and refine the candidate words:

针对用户输入的数据进行分词并去除停用词，将剩余的分词作为候选词；Segment the data entered by the user and remove stop words, and use the remaining word segmentation as candidate words;

Step3、根据候选词的个数，计算距离并排序输出：Step3. According to the number of candidate words, calculate the distance and sort the output:

Step3.1、若候选词个数为1时：Step3.1, if the number of candidate words is 1:

从关键字表B中获取与候选词的首字x联结的字、首字x之间的特征值Aix、获取尾字y、与尾字y联结的字之间的特征值Ayj；计算Aix≠0情况下首字与关键字表B中字的距离dix且得到ixy对应的词组合，计算Ayj≠0情况下尾字与关键字表B中字的距离dyj且得到xyj对应的词组合；根据dix、dyj从小到大的顺序排列其对应的词组合；根据词组合的顺序，将词组合与关键词表A中的关键词进行匹配获取对应的法规名称，将匹配的结果去除重复后按照顺序显示；其中，当出现dix＝dyj，则dix、dyj对应的词组合进行随机排序；Obtain from keyword table B the word that is connected with the first word x of the candidate word, the characteristic value Aix between the first word x, obtain the last word y, and the characteristic value Ayj between the words connected with the last word y; calculate Aix≠ In the case of 0, the distance dix between the first character and the word in the keyword table B is obtained and the word combination corresponding to ixy is obtained, and the distance dyj between the last word and the word in the keyword table B is calculated under the situation of Ayj ≠ 0 and the word combination corresponding to xyj is obtained; according to dix, dyj arranges the corresponding word combinations in order from small to large; according to the order of the word combinations, match the word combinations with the keywords in the keyword table A to obtain the corresponding legal names, remove the duplicates of the matching results and display them in order; Wherein, when dix=dyj occurs, then the word combinations corresponding to dix and dyj are randomly sorted;

Step3.2、若候选词个数不为1时：Step3.2, if the number of candidate words is not 1:

将多个候选词按输入顺序排列，分别计算相邻两个候选词中先输入的候选词的尾字u与后输入的候选词的首字v的距离duv及对应的两个候选词构成的词组合；从关键字表B中获取与各个候选词的首字x联结的字、首字x之间的特征值Aix、获取尾字y、与尾字y联结的字之间的特征值Ayj；计算Aix≠0情况下首字与关键字表B中字的距离dix且得到ixy对应的词组合，计算Ayj≠0情况下尾字与关键字表B中字的距离dyj且得到xyj对应的词组合；根据duv、dix、dyj从小到大的顺序排列其对应的词组合；根据词组合的顺序，将词组合与关键词表A中的关键词进行匹配获取对应的法规名称，将匹配的结果去除重复后按照顺序显示；其中，当出现duv＝dix＝dyj，则仅仅保留duv对应的词组合进行排序，当出现dix＝dyj，则dix、dyj对应的词组合进行随机排序；Arrange a plurality of candidate words in the input order, respectively calculate the distance duv between the last word u of the candidate word input first and the first word v of the candidate word input later among the two adjacent candidate words and the corresponding two candidate words Word combination; from the keyword table B, the character value Aix between the first word x and the first word x of each candidate word is obtained, and the end word y is obtained, and the characteristic value Ayj between the words connected with the last word y ; Calculate the distance dix between the first word and the word in the keyword table B under the condition of Aix≠0 and obtain the word combination corresponding to ixy, calculate the distance dyj between the last word and the word in the keyword table B under the situation of Ayj≠0 and obtain the word combination corresponding to xyj ; According to the order of duv, dix, dyj from small to large, arrange their corresponding word combinations; according to the order of word combinations, match the word combinations with the keywords in keyword table A to obtain the corresponding legal name, and remove the matching result Display in order after repetition; wherein, when duv=dix=dyj occurs, only the word combinations corresponding to duv are reserved for sorting, and when dix=dyj occurs, then the word combinations corresponding to dix and dyj are randomly sorted;

所述u、v、x、y为字在关键字表B中的标记。The u, v, x, and y are marks of words in the keyword table B.

实施例3：如图1-2所示，一种用于环境保护法规检索的关联优先排序方法，首先对环保法律法规检索系统构建一个关键词表A、关键字表B；然后清洗用户输入的数据并提炼候选词；最后根据候选词的个数，计算距离并排序输出。Embodiment 3: As shown in Fig. 1-2, a kind of association prioritization method that is used for the retrieval of environmental protection laws and regulations, first builds a keyword table A, keyword table B to the environmental protection laws and regulations retrieval system; Then cleans the user input Data and extract candidate words; finally, according to the number of candidate words, calculate the distance and sort the output.

实施例4：如图1-2所示，一种用于环境保护法规检索的关联优先排序方法，首先对环保法律法规检索系统构建一个关键词表A、关键字表B；然后清洗用户输入的数据并提炼候选词；最后根据候选词的个数，计算距离并排序输出。Embodiment 4: As shown in Fig. 1-2, a kind of association prioritization method that is used for the retrieval of environmental protection laws and regulations, first builds a keyword table A, keyword table B to the environmental protection laws and regulations retrieval system; Then clean the user input Data and extract candidate words; finally, according to the number of candidate words, calculate the distance and sort the output.

所述方法的具体步骤如下：The concrete steps of described method are as follows:

对环保法律法规检索系统构建一个关键词表A、关键字表B；其中，关键词表A：存储着法规名称及法规中抽取出来的t组关键词；关键字表B：存储着关键词表A中每个关键词拆分成的不同字m个及各个字之间的特征值Aij；Aij表示角标为i和j所代表的字的组合出现在关键词表A中的频数，角标i、j为关键词表A中每个关键词拆分成的不同字在关键字表B中的标记；Construct a keyword table A and keyword table B for the retrieval system of environmental protection laws and regulations; among them, keyword table A: store the name of the law and t groups of keywords extracted from the law; keyword table B: store the keyword table Each keyword in A is divided into m different words and the characteristic value Aij between each word; Aij indicates the frequency of the combination of words represented by subscripts i and j appearing in keyword table A, and subscripts i, j are the marks in the keyword table B of the different words that each keyword is split into in the keyword table A;

表A-关键词表Table A - Keyword Table

关键字表B：存储着关键词表A中每个关键词拆分成的不同字(称之为元素)m个及各个元素之间的特征值Aij，Aij表示角标为i和j所代表的字的组合出现在关键词表A中的频数。Keyword table B: Stores m different words (called elements) divided into each keyword in keyword table A and the characteristic value Aij between each element, Aij means that the subscripts are represented by i and j The frequency of the combination of words appearing in the keyword list A.

表B-关键字表Table B - Keyword Table

从表A中，可以看出，t＝40；从表B中，可知，i＝j＝1，2，...,40。From Table A, it can be seen that t=40; from Table B, it can be seen that i=j=1, 2, . . . , 40.

Step2、清洗数据并提炼候选词：Step2, clean the data and refine the candidate words:

查询的本质，就是对词语词组的匹配或查询，则需要获取能表示替代用户查询意图的词语或词组，为方便表示，我们称之为候选词。The essence of query is the matching or query of words and phrases. It is necessary to obtain words or phrases that can represent the user's query intention. For the convenience of expression, we call them candidate words.

Step2.1、假设用户输入“污染”，则：Step2.1, assuming the user input "pollution", then:

清洗用户输入的数据，将其分词并去除停用词，抽取出候选词；即提炼出候选词——“污染”一个候选词。Clean the data entered by the user, segment it into words and remove stop words, and extract candidate words; that is, extract candidate words—"pollution" a candidate word.

Step3、判断候选词类型并计算距离并排序输出：Step3. Determine the candidate word type and calculate the distance and sort the output:

判断候选词个数为一，则计算候选词“污染”的首字与尾字与关键字表中各个字的距离；读取关键字表，易知：与首字“污”联结的字与其之间的特征值如下：A12＝3，A22＝1，A32＝1，A52＝3，A62＝1，A72＝1，A82＝3；与尾字“染”联结的字与其之间的特征值如下：A94＝2，A95＝6，A96＝6；在这里，联结方式为：若是找到与词语的首字的联结组合，则找到形为“**+首字”的组合，若是找到与词语的首字的联结组合，则找到形为“尾字+**”的组合，下文不再赘述；Judging that the number of candidate words is one, then calculate the distance between the first word and the last word of the candidate word "pollution" and each word in the keyword table; read the keyword table, it is easy to know: the word connected with the first word "pollution" The characteristic value between is as follows: A12=3, A22=1, A32=1, A52=3, A62=1, A72=1, A82=3; The word that links with tail word " dyeing " and characteristic value between them As follows: A94=2, A95=6, A96=6; Here, the connection mode is: if find the connection combination with the first word of the word, then find the combination that is shaped as "**+first word", if find the combination with the first word of the word The connection combination of the first word of the first word, then find the combination of the form "tail word+**", which will not be described in detail below;

则本实例中该词的第一个字与关键字表中字的距离为：Then the distance between the first character of the word and the word in the keyword table in this example is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>12</span>12</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>12</span>12</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>3</span>3</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>22</span>twenty\; two</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>22</span>twenty\; two</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>32</span>32</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>32</span>32</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>52</span>52</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>52</span>52</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>3</span>3</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>62</span>62</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>62</span>62</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>72</span>72</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>72</span>72</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>82</span>82</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>82</span>82</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>3</span>3</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

本实例中该词的第二个字与关键字表中字的距离为：The distance between the second character of this word and the word in the keyword table in this example is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>94</span>94</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>94</span>94</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>2</span>2</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>95</span>95</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>95</span>95</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>6</span>6</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>96</span>96</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>96</span>96</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>6</span>6</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

根据字之间距离越小相关性越大，将距离d从小到大排列，若值相等则随机排列，则其顺序为：d95,d96,d12,d52,d82,d94,d22,d32,d62,d72；将关键字表里面字的与输入词中的字组合起来，回到关键词表中去匹配：According to the smaller the distance between words, the greater the correlation, arrange the distance d from small to large, if the values are equal, arrange randomly, the order is: d95, d96, d12, d52, d82, d94, d22, d32, d62, d72; combine the words in the keyword table with the words in the input word, and return to the keyword table to match:

首先，由以上实例计算结果按距离从小到大(距离值并列则随机排列)可以得到的组合为：“污染物”、“污染防”、“水污染”、“声污染”、“气污染”、“污染源”、“源污染”、“放污染”、“活污染”、“治污染”。First of all, according to the calculation results of the above example, the combinations that can be obtained according to the distance from small to large (the distance values are juxtaposed are randomly arranged): "pollutant", "pollution prevention", "water pollution", "sound pollution", "air pollution" , "pollution source", "source pollution", "release pollution", "live pollution", "pollution control".

然后，将得到的上列组合与关键词表A中的关键词匹配，看其是否存在于关键词表中，若存在，则将该关键词所对应的法规优先显示输出，若匹配不存在则进行下一组合的匹配。根据上列组合：“污染物”能够和关键词表A中的“污染物”、“大气污染物”、“污染物排放”、“陆源污染物”匹配，可以索引到如下法规：Then, match the obtained above combination with the keyword in keyword table A to see if it exists in the keyword table, if it exists, then the regulations corresponding to the keyword will be displayed and output first, if the match does not exist then Match the next combination. According to the above combination: "pollutant" can be matched with "pollutant", "atmospheric pollutant", "pollutant discharge" and "land-based pollutant" in keyword table A, and can be indexed to the following regulations:

《中华人民共和国水污染防治法》"Water Pollution Prevention and Control Law of the People's Republic of China"

《中华人民共和国大气污染防治法》"Law of the People's Republic of China on the Prevention and Control of Air Pollution"

《中华人民共和国海洋环境保护法》"Marine Environmental Protection Law of the People's Republic of China"

《中华人民共和国环境保护法》"Environmental Protection Law of the People's Republic of China"

所以将这些法规规优先显示，后序组合中如：“污染防”能够和关键词表A中的“水污染防治”、“大气污染防治”、“噪声污染防治”、“污染防治”匹配，可以索引到如下法规：Therefore, these laws and regulations are displayed first. In the subsequent combination, for example: "pollution prevention" can match "water pollution prevention", "air pollution prevention", "noise pollution prevention" and "pollution prevention" in keyword list A. The following regulations can be indexed:

《中华人民共和国水污染防治法》、《中华人民共和国大气污染防治法》、《中华人民共和国环境噪声污染防治法》、《饮用水水源保护区污染防治管理规定》、《中华人民共和国水污染防治法实施细则》，去重复，将本条得到的法规与之前组合索引得到的法规去重复，得到如下法规：《中华人民共和国环境噪声污染防治法》、《饮用水水源保护区污染防治管理规定》、《中华人民共和国水污染防治法实施细则》。"Water Pollution Prevention and Control Law of the People's Republic of China", "Air Pollution Prevention and Control Law of the People's Republic of China", "Environmental Noise Pollution Prevention and Control Law of the Detailed Rules for the Implementation of the Law, deduplication, deduplication of the regulations obtained in this article and the regulations obtained in the previous combination index, and the following regulations are obtained: "Law of the People's Republic of China on the Prevention and Control of Environmental Noise Pollution", "Provisions on the Prevention and Control of Pollution in Drinking Water Source Protection Areas", Detailed Rules for the Implementation of the Water Pollution Prevention and Control Law of the People's Republic of China.

后序组合依次类推。Subsequent combinations and so on.

Step4、假设用户输入“污染与防治”，则：Step4. Assuming that the user inputs "pollution and prevention", then:

清洗用户输入的数据，将其分词并去除停用词，抽取出候选词；即分词，去除停用词“与”提炼出候选词——“污染”和“防治”两个候选词。Clean the data entered by the user, segment it and remove stop words, and extract candidate words; that is, word segmentation, remove stop words "and" and extract candidate words—two candidate words "pollution" and "control".

判断候选词类型并计算距离并排序输出：Determine the type of candidate words and calculate the distance and sort the output:

判断候选词个数不为一，则：Judging that the number of candidate words is not one, then:

首先计算候选词“污染”和“防治”这两个候选词之间的距离，即计算出“染”与“防”这两个字的距离。读取关键字表B，易知：“染”与“防”这两个关键字的特征值为A96＝6；则本实例中这两个候选词之间的距离为：First, the distance between the two candidate words "pollution" and "control" is calculated, that is, the distance between the two words "dye" and "defense" is calculated. Read the keyword table B, it is easy to know: the characteristic value of the two keywords of "dye" and "defense" is A96=6; then the distance between these two candidate words in this example is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>96</span>96</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>96</span>96</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>6</span>6</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

然后，计算各个候选词的首字与尾字与关键字表中各个字的距离；读取关键字表，易知：Then, calculate the distance between the first word and the last word of each candidate word and each word in the keyword table; read the keyword table, it is easy to know:

与第一个候选词“污染”的首字“污”联结的字与其之间的特征值如下：A12＝3，A22＝1，A32＝1，A52＝3，A62＝1，A72＝1，A82＝3；与尾字“染”联结的字与其之间的特征值如下：A94＝2，A95＝6，A96＝6；The character value between the word that is connected with the first word "pollution" of the first candidate word "pollution" is as follows: A12=3, A22=1, A32=1, A52=3, A62=1, A72=1, A82=3; Characteristic value between the word that is connected with suffix " dye " and it is as follows: A94=2, A95=6, A96=6;

则本实例中该词的第一个字与关键字表中字的距离为：Then the distance between the first character of the word and the word in the keyword table in this example is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>12</span>12</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>12</span>12</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>3</span>3</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>22</span>twenty\; two</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>22</span>twenty\; two</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>32</span>32</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>32</span>32</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>52</span>52</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>52</span>52</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>3</span>3</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>62</span>62</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>62</span>62</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>72</span>72</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>72</span>72</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>82</span>82</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>82</span>82</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>3</span>3</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

本实例中该词的第二个字与关键字表中字的距离为：The distance between the second character of this word and the word in the keyword table in this example is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>94</span>94</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>94</span>94</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>2</span>2</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>95</span>95</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>95</span>95</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>6</span>6</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>96</span>96</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>96</span>96</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>6</span>6</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

计算第二个候选词“防治”的首字与尾字和关键字表中各个字的距离；读取关键字表B，易知：与首字“防”联结的字与其之间的特征值如下：A46＝1，A96＝6与尾字“治”联结的字与其之间的特征值皆为A72＝1；Calculate the distance between the first word and the last word of the second candidate word "prevention" and each word in the keyword table; read the keyword table B, it is easy to know: the character connected with the first word "defense" and the eigenvalues between them As follows: A46=1, A96=6 is A72=1 with the word that A96=6 is connected with tail word " governance " and the feature value between them;

则本实例中“防治”这一词的首字与关键字表中的距离为：Then the distance between the first character of the word "control" and the keyword table in this example is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>46</span>46</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>46</span>46</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>96</span>96</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>96</span>96</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>6</span>6</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

本实例中该词尾字与关键字表中字的距离为：In this example, the distance between the suffix word and the word in the keyword table is:

$\mathrm{<span><span>d</span>d</span>}\mathrm{<span><span>72</span>72</span>}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>A</span>A</span>}\mathrm{<span><span>72</span>72</span>}}}{\mathrm{<span><span>t</span>t</span>}}<span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}<span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}^{<span><span>-</span>-</span>\mathrm{<span><span>1</span>1</span>}}}{\mathrm{<span><span>40</span>40</span>}}<span><span>;</span>;</span>$

根据字之间距离越小相关性越大，将距离d从小到大排列，若值相等则随机排列，其顺序为：d96,d95,d12,d52,d82,d94,d22,d32,d62,d72,d46；将关键字表里面字的与输入词中的字组合起来，回到关键词表中去匹配：According to the smaller the distance between words, the greater the correlation, arrange the distance d from small to large, if the values are equal, arrange randomly, the order is: d96, d95, d12, d52, d82, d94, d22, d32, d62, d72 ,d46; combine the words in the keyword table with the words in the input word, and return to the keyword table to match:

首先，由以上实例计算结果按距离从小到大(距离值并列则随机排列)可以得到的组合为：“污染防治”、“污染物”、“水污染”、“声污染”、“气污染”、“污染源”“源污染”、“放污染”、“活污染”、“治污染”、“境防治”；(“污染防治”为计算两个候选词之间的距离所得到的组合，由于两个候选词之间的距离所得到的d96与后续的词的首尾字与关键字表中的字的组合之间的距离出现同一个值，即d96，所以看起来有问题，实际上当两个词之间的距离与其他字的组合的距离出现重复距离值时，选择两个词之间的距离组合)First of all, according to the calculation results of the above example, the combinations that can be obtained according to the distance from small to large (the distance values are arranged in parallel) are: "pollution prevention", "pollutant", "water pollution", "sound pollution", "air pollution" , "pollution source", "source pollution", "release pollution", "live pollution", "pollution control", "environmental control"; ("pollution control" is the combination obtained by calculating the distance between two candidate words, because The d96 obtained by the distance between two candidate words and the distance between the combination of the first and last words of the subsequent words and the words in the keyword table have the same value, that is, d96, so it seems that there is a problem. In fact, when two The distance between words and the distance between other word combinations When there are repeated distance values, select the distance combination between two words)

然后，将得到的上列组合与关键词表A中的关键词匹配，看其是否存在于关键词表中，若存在，则将该关键词所对应的法规优先显示输出，若匹配不存在则进行下一组合的匹配。Then, match the obtained above combination with the keyword in the keyword table A to see if it exists in the keyword table. If it exists, the regulations corresponding to the keyword will be displayed and output first. If the match does not exist, then Match the next combination.

根据上列组合：According to the combination of the above:

“污染防治”能够和关键词表A中的“水污染防治”、“大气污染防治”、“噪声污染防治”、“污染防治”匹配，索引得到如下法规结果："Pollution prevention and control" can be matched with "water pollution prevention and control", "air pollution prevention and control", "noise pollution prevention and control" and "pollution prevention and control" in Keyword Table A, and the index results are as follows:

《中华人民共和国水污染防治法》"Water Pollution Prevention and Control Law of the People's Republic of China"

《中华人民共和国大气污染防治法》"Law of the People's Republic of China on Prevention and Control of Air Pollution"

《中华人民共和国环境噪声污染防治法》"Law of the People's Republic of China on the Prevention and Control of Environmental Noise Pollution"

《饮用水水源保护区污染防治管理规定》Regulations on the Prevention and Control of Pollution in Drinking Water Source Protection Areas

《中华人民共和国水污染防治法实施细则》Detailed Rules for the Implementation of the Water Pollution Prevention and Control Law of the People's Republic of China

“境防治”能够和关键词表A中的“水污染防治”、“大气污染防治”匹配，索引得到如下法规："Environmental prevention and control" can be matched with "water pollution prevention and control" and "air pollution prevention and control" in Keyword Table A, and the following regulations are indexed:

《中华人民共和国固体废物污染环境防治法》；"Law of the People's Republic of China on the Prevention and Control of Environmental Pollution by Solid Waste";

所以这些法规优先显示，其他组合依次类推。Therefore, these regulations are displayed first, and other combinations are deduced by analogy.

上面结合附图对本发明的具体实施方式作了详细说明，但是本发明并不限于上述实施方式，在本领域普通技术人员所具备的知识范围内，还可以在不脱离本发明宗旨的前提下作出各种变化。The specific implementation of the present invention has been described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned implementation, within the knowledge of those of ordinary skill in the art, it can also be made without departing from the gist of the present invention. Variations.

Claims (3)

1., for an association priority ordering method for legislations of environmental protection retrieval, it is characterized in that: first an antistop list A, key table B are built to environmental law Laws & Regulations searching system; Then clean the data of user's input and refine candidate word; Finally according to the number of candidate word, calculate distance and output of sorting.

2. the association priority ordering method for legislations of environmental protection retrieval according to claim 1, is characterized in that: the concrete steps of the described association priority ordering method for legislations of environmental protection retrieval are as follows:

Step1, first set up system model:

An antistop list A, key table B are built to environmental law Laws & Regulations searching system; Wherein, antistop list A: store the t group keyword extracted in regulation title and regulation; Key table B: store the eigenwert Aij between different word m and each word that each keyword in antistop list A splits into; Aij represents that the combinatorics on words of footmark representated by i and j appears at the frequency in antistop list A, and footmark i, j are the mark of different words in key table B that in antistop list A, each keyword splits into;

The data that Step2, cleaning user input also refine candidate word:

Data for user's input are carried out participle and remove stop words, by remaining participle alternatively word;

Step3, number according to candidate word, calculate distance and output of sorting:

If when Step3.1 candidate word number is 1:

Obtain from key table B and word that the lead-in x of candidate word connects, eigenwert Aix between lead-in x, obtain eigenwert Ayj between word that tail word y and tail word y connects; Calculate the distance dix of lead-in and word in key table B in Aix ≠ 0 situation and obtain word combination corresponding to ixy, the distance dyj of calculating Ayj ≠ 0 situation lower tail word and word in key table B and obtain word combination corresponding to xyj; The word combination of its correspondence is arranged according to dix, dyj order from small to large; According to the order of word combination, word combination and the keyword in antistop list A are carried out mating obtain corresponding regulation title, the result of coupling is removed after repeating and show in order; Wherein, when there is dix=dyj, then the word combination that dix, dyj are corresponding carries out randomly ordered;

If when Step3.2 candidate word number is not 1:

Multiple candidate word is pressed input sequence arrangement, calculate the tail word u of the candidate word first inputted in adjacent two candidate word and distance duv and corresponding two word combination that candidate word is formed of the lead-in v of the candidate word of rear input respectively; Obtain from key table B and word that the lead-in x of each candidate word connects, eigenwert Aix between lead-in x, obtain eigenwert Ayj between word that tail word y and tail word y connects; Calculate the distance dix of lead-in and word in key table B in Aix ≠ 0 situation and obtain word combination corresponding to ixy, the distance dyj of calculating Ayj ≠ 0 situation lower tail word and word in key table B and obtain word combination corresponding to xyj; The word combination of its correspondence is arranged according to duv, dix, dyj order from small to large; According to the order of word combination, word combination and the keyword in antistop list A are carried out mating obtain corresponding regulation title, the result of coupling is removed after repeating and show in order; Wherein, when there is duv=dix=dyj, then the word combination only retaining duv corresponding sorts, and when there is dix=dyj, then the word combination that dix, dyj are corresponding carries out randomly ordered;

Described u, v, x, y are the mark of word in key table B.

3. the association priority ordering method for legislations of environmental protection retrieval according to claim 2, is characterized in that: described in$duv=\frac{1+e^{-Auv}}{t},dix=\frac{1+e^{-Aix}}{t},dyj=\frac{1+e^{-Ayj}}{t};$Wherein Auv, Aix, Ayj represent the combinatorics on words of footmark representated by u, v respectively, the combinatorics on words of footmark representated by i, x, and the combinatorics on words of footmark representated by y, j appears at the frequency in antistop list A; Duv, dix, dyj represent the word of footmark representated by u, v respectively, the word of footmark representated by i, x, the distance of the word of footmark representated by y, j.