CN103646017B - Acronym generating system for naming and working method thereof - Google Patents

Abstract

The invention discloses an abbreviation generating system for naming and a working method thereof. The system gives appropriate abbreviation names by analyzing input long character strings. The system includes input and output pages and a background support service platform. The input and output pages are used to input long strings of acronyms to be generated and output acronyms for naming; the background support service platform includes a classification database and an acronym generation system And the recommendation system, the classification database stores data that is convenient for calculating the classification tendency of each word, which can make the background support program analyze the sentences entered by the user and generate semantically related acronyms for the naming of various fields. The purpose of the present invention is to change the status quo, so that users can obtain abbreviations that are semantically similar to the original character strings for naming.

Description

Acronym Generation System for Naming and How It Works

Technical field:

The present invention relates to a name generation system and its working method, in particular to an acronym generation system and its working method for naming, which provides a string composed of multiple words with an abbreviation that matches the meaning of the string itself. Abbreviations.

Background technique:

Acronym generation technology is a creative automatic generation technology that simulates human thinking to analyze character strings. Since every industry and field needs to create a well-known good brand, it is particularly important to choose a name that is catchy and professional. Therefore, building an acronym generation system for naming has great application potential and broad prospects in various industries and fields.

At present, there is an acronym generation and acronym query system on the Internet, and its naming rules basically use initial letter matching, that is, the first letter of all words in the string is extracted to form an acronym name. However, there is no feasible solution for how to generate acronyms related to the meaning of strings. Some acronym generation systems rank the generated acronyms by voting by users, but because voting requires the cooperation of users, it is not satisfactory for strings with a small number of queries. Moreover, considering the different needs of users in different fields, the performance of the acronyms generated by the voting method in the semantics of the acronyms is not satisfactory.

In the existing acronym generation system for naming, we use computers to analyze the semantics of character strings, and construct acronyms with similar semantics through correlation matching rules. Check whether such a combination constitutes a word by querying the background database, and if not, add or delete some letters to achieve the purpose of generating an acronym.

Invention content:

The present invention provides an abbreviation generating system for naming and its working method. By changing the existing abbreviation generating system, it can generate abbreviations matching the semantics of the original character string.

The present invention adopts the following technical solutions: an acronym generation system for naming, the system includes an input and output page and a background support service platform; wherein, the input and output page is used for inputting a long string of acronyms to be generated and outputting Acronyms used for naming; background support service platforms include:

Classification database, the classification database is established using background support programs, and the classification database stores the data that calculates the classification tendency of each word, which is used to find the classification of the sentence entered by the user, and is used to match the required abbreviation in the database to which the corresponding classification belongs. Abbreviations;

The acronym generation system analyzes the semantics and category of the original sentence by querying the classification database, so as to find out the acronyms that match the semantics in the subsequence of the sentence string and sort them according to the degree of semantic relevance;

In the recommendation system, when the acronyms do not match, it can judge the semantics of the input sentence, and modify some words in the sentence or exchange the order of the words in the sentence without affecting the semantics of the sentence, and then compare it with the words in the classification database Matching is performed so that it can generate matching acronyms and recommend them to users.

The present invention also adopts following technical scheme: a kind of working method for the acronym generating system of naming, it comprises the steps:

1). Enter the long character string of the acronym to be generated, and confirm the generation;

2). According to the long string entered above, extract each word in the string, match it with the words in the classification database in turn, and calculate the type of each word;

3). Determine the type of the string, and then save the type;

4). Analyze the first letter of the real word in the long string as a fixed letter, and on this basis, keep the original order of the characters and insert other letters in the long string, find out all possible acronyms, and follow the steps in turn 3) match the word in the database corresponding to the type found in, if the match is successful, then save it as a candidate acronym;

5). Sort all the candidate acronyms according to the degree of type correlation, and the degree of type correlation is obtained in the type database;

6). Display the sorted acronyms in the acronym output box, go to step 7), if the relevant type of acronym cannot be generated, go to step 8);

7). Perform a reset action to prepare for the next generation of acronyms;

8). Enter the acronym recommendation system, modify the sentence without changing the semantics of the sentence, generate acronyms, and feed back the modified sentence to the user.

The present invention has following beneficial effect:

(1) The present invention calculates the classification of the input character string through the classification of words, and matches each subsequence of the input character string under the classification, so as to achieve the purpose that the semantics of the acronym is similar to the input character string;

(2) Compared with the existing acronym generation system, the present invention can greatly improve the degree of correlation between the generated named acronym and the original sentence.

Description of drawings:

FIG. 1 is a structural diagram of an acronym generation system for naming in the present invention.

Fig. 2 is a flow chart of abbreviation generation in the present invention.

Fig. 3 is a flow chart for generating the classification database of the present invention.

detailed description:

Please refer to FIG. 1 to FIG. 3 , the acronym generation system for naming in the present invention includes input and output pages and background support programs. The input and output pages include a statement input box, a naming generation button, a reset button, a recommendation button, and an acronym output box. The background support program is divided into three parts: the generation of classification database, the generation of acronyms and the recommendation system. The taxonomy database is used to find the category to which the sentence entered by the user belongs, and to match the required acronym in the database to which the corresponding category belongs. These three parts are described in detail below:

(1) Generation of classification database

The classification database of the present invention is established in advance using a background support program, and the stored data includes a large number of words and the classification tendencies of each word. Building a classification database requires a large amount of training text. We first count all the words that appear in the training text, then calculate the importance of each word to each independent text, and finally use the principle of cosine similarity to classify the text to obtain the word classification. Generate a classification database through the training text, the steps are as follows:

A: Analyze the original text file by the processing program to obtain the number of occurrences of each word in each text, and store it in the format of <word, file ID [number of occurrences]> in the preliminary database;

B: Calculate the importance ti of the word to each file for each tuple in the preliminary database. Here, TF-IDF (term frequency-inverse document frequency) technology is used:

${\mathrm{<span\; class="notranslate">\; tf</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}}$

${\mathrm{<span\; class="notranslate">\; idf</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; log</span>}\frac{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; :</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; |</span>}$

tndf _{i, j} = tf _{i, j} × idf _i

In the above formula, n _{i, j} is the number of occurrences of the word in file d _j , and the denominator is the sum of the number of occurrences of all words in file d _j . |D| indicates the total number of documents, and |{j:t _i ∈ d _j }| indicates the number of documents containing the word ti.

In this way, the degree of correlation between all words and files in the preliminary database is calculated and stored in the define database. The format is <word, degree of correlation with file 1, degree of correlation with file 2, ..., degree of correlation with file n>.

C: The degree of correlation between each word in the define database described in B and file i forms a vector, and the cosine similarity between files is calculated:

$\mathrm{<span\; class="notranslate">\; similarity</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; cos</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; B</span>}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\sqrt{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \&times;</span>\sqrt{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}$

D: Classify files according to the cosine similarity calculated in C, and then classify each word according to the classification of files, so that the degree of correlation between each word and a certain type is obtained, and a classification database is formed.

(2) Generation of acronyms

Acronyms are generated in two stages:

(A): Analyze the semantics of user input strings:

Firstly, extract each word in the sentence input by the user, and analyze the category of each word by querying the classification database established in (1), so as to obtain and record the category of the sentence.

(B): Generate semantically related acronyms and sort them:

Acronym matching:

Considering that good acronyms often use the initials in the original sentence as much as possible, the present invention first extracts the initials of the user input sentence to form an acronym and matches the database under the classification recorded in (A), if the matching is successful, This acronym is recorded as the best acronym.

Insert other word matches in the sentence based on the first letter:

The odds of a successful acronym match are not very high. When the matching fails, we consider selecting some words in the original sentence, inserting them into the initial sequence in the original order, and matching again in the database under the classification recorded in (A).

Among them, the acronym generation part is initiated by the user, input a string containing N words, click the generate button, and hand it over to the background processing program, the steps are as follows:

A. Extract all the words in the input string to form a sequence a, and extract all the characters in the input string to form a sequence b;

B. According to the sequence a described in step A, use the TF-IDF technology mentioned in the preparatory part to calculate the classification of the original character string;

C. according to classification in step B, all words under this type database are taken out, form a Trie tree, so that the matching of abbreviations;

D. For the sequence b described in step A, according to the requirements in step 4) of the summary of the invention, traverse all subsequences whose length is less than or equal to N+2, and search for the subsequence in the Trie tree described in step C, if Find and save all results to sequence c;

E. Sort and print the sequence c to the screen according to the degree to which it is associated with the type in the taxonomy database. If the relevant type of acronym cannot be generated, or a user request is received, go to step G;

F. Click the reset button to prepare for the next generation of acronyms;

G. Enter the acronym recommendation system, modify the sentence without changing the semantics of the sentence, generate an acronym, and feed back the modified sentence to the user.

(3) Recommendation system

If the user is dissatisfied with the acronyms generated in (2), a better acronym will be provided by modifying the original sentence through the recommendation system. The recommendation system can guarantee that the sentence can be modified without changing the meaning of the original sentence. After modifying the sentence, match it again with the words in the classification database recorded in (2). The following are three feasible methods for modifying sentences in this system:

A. Reorder the words in the original sentence

This method can simply change the order of the words in the original sentence, which can greatly guarantee the integrity of the meaning of the original sentence. For example: extract the first letter of the original sentence, and exchange the order: in order to retain the meaning of the original sentence and ensure the correct grammar, you can consider exchanging the words connected by and or or.

B. Insert synonyms into the original sentence

When using this method, you can consider using conjunctions such as and or or between synonyms to ensure grammatical correctness.

C. Modify synonyms or words of the same type in the original sentence

Use existing dictionaries to find synonymous words and replace content words in long character strings.

Please refer to shown in Fig. 1 to Fig. 3, the working method of the acronym generation system that the present invention is used for naming comprises the steps:

1). The user enters a long string of acronyms to be generated and clicks the generate button;

2). The background processing program extracts each word in the string according to the user's input, and sequentially matches with the words in the classification database and calculates the type of each word;

3). The background processing program determines the most likely type of the string, and then saves the type;

4). The background processing program finds out all the subsequences of the long string, and matches them with the words in the database corresponding to the type found in step 3). If the match is successful, it is saved as a candidate acronym;

5). Sort all the candidate acronyms according to the degree of type correlation, which can be obtained in the type database;

6). Display the sorted acronyms in the acronym output box, go to step 7), if the relevant type of acronym cannot be generated, go to step 8);

7). Click the reset button to prepare for the next generation of acronyms;

8). Enter the acronym recommendation system, modify the sentence without changing the semantics of the sentence, generate acronyms, and feed back the modified sentence to the user.

The system for generating abbreviations used for naming and the working method thereof of the present invention change the existing system for generating abbreviations so that they can generate abbreviations that match the semantics of the original character strings.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, some improvements can also be made without departing from the principle of the present invention, and these improvements should also be regarded as the invention. protected range.

Claims (2)

1. an acronym generation system for naming, it is characterized in that, the system includes input and output pages and background support service platform; The acronym named after; the background support service platform includes: Classification database, the classification database is established using background support programs, and the classification database stores the data that calculates the classification tendency of each word, which is used to find the classification of the sentence entered by the user, and is used to match the required abbreviation in the database to which the corresponding classification belongs. Abbreviations; The acronym generation system analyzes the semantics and category of the original sentence by querying the classification database, so as to find out the acronyms that match the semantics in the subsequence of the sentence string and sort them according to the degree of semantic relevance; In the recommendation system, when the acronyms do not match, by judging the semantics of the input sentence, modifying some words in the sentence or changing the order of the words in the sentence without affecting the semantics of the sentence, and then matching the words in the classification database Matching is performed so that it can generate matching acronyms and recommend them to users. 2. A working method for the abbreviation generation system for naming as claimed in claim 1, is characterized in that comprising the steps: 1). Enter the long character string of the acronym to be generated, and confirm the generation; 2). According to the long character string input above, extract each word in the character string, and sequentially match the words in the classification database and calculate the type of each word; 3). Determine the type of the string, and then save the type; 4). Analyze the first letter of the real word in the long string as a fixed letter, and on this basis, keep the original order of the characters and insert other letters in the long string, find out all possible abbreviations, and follow the steps in turn 3) match the word in the database corresponding to the type found in, if the match is successful, then save it as a candidate acronym; 5). Sorting all candidate acronyms according to the degree of type correlation, and the degree of type correlation is obtained in the type database; 6). Display the sorted acronyms in the acronym output box, go to step 7), if the relevant type of acronym cannot be generated, go to step 8); 7). Perform a reset action to prepare for the next generation of acronyms; 8). Enter the acronym recommendation system, modify the sentence without changing the semantics of the sentence, generate an acronym, and feed back the modified sentence to the user.