CN106168946A - A kind of method identifying user initials phenomenon - Google Patents

Abstract

The invention provides a method for identifying the abbreviation phenomenon of user names. The steps include: 1) filtering the characters in two or more user names, and only keeping English letters and numbers; 2) filtering the above-mentioned filtered user names Each is divided into several consecutive segments, and the first character of each segment is selected to form a new character string; 3) The length of the longest abbreviation is obtained according to the above new character string. If the length value is greater than or equal to a given threshold ΔL, then the There are abbreviations of user names among the above user names; the reserved English letters are uniformly converted into lowercase or uppercase; the fragments are words or single characters; the fragments are segmented according to the specified dictionary; new string gets the longest abbreviation length.

Description

A Method for Recognizing Username Abbreviation Phenomenon

technical field

The invention relates to the field of computers, in particular to a method for identifying abbreviated user names.

Background technique

In recent years, the Internet has developed rapidly and has penetrated into all aspects of social life, such as watching news or videos on portal sites such as Sina, Sohu, and Tencent, and exchanging information on social networks such as Weibo, Tieba, and communities. When people use these networks Will register an account, fill in the user name. The user name is a character string that meets certain rules and can identify the user's identity filled in by the user when registering for the website. It usually consists of English letters, numbers, and underscores and other special characters.

When a user registers on a certain website, because the user name of the website is unique, and the commonly used user name has been registered by others, or due to other considerations such as protecting personal privacy, the user does not use the commonly used user name, but at the same time, for the sake of easy memory And choose to abbreviate some words in the commonly used user name, become a new user name and register for use, this is the abbreviation phenomenon of the user name. Identifying abbreviations between user names is of great significance to Internet research. For example, when researchers are mining social network data, such as user behavior analysis, personalized recommendation, etc., sometimes it is necessary to analyze the same user or similar users in different social networks. Association, and some of the association methods need to analyze the similarity of user names by identifying abbreviations in user names.

Recognizing the abbreviation phenomenon of usernames is to give two usernames and judge whether a certain fragment of one username is an abbreviation of a certain fragment of the other username. The general method to solve this problem is the enumeration method, that is, to enumerate the substring s ₁ of a user name and the substring s ₂ of another user name, and check whether s ₁ is an abbreviated form of s ₂ . If such s ₁ and s ₂ exist, it is considered that there is an abbreviation between the two user names, and s ₁ is an abbreviation of s ₂ , otherwise it is considered that there is no abbreviation between the two user names.

However, the enumeration method needs to enumerate all the substrings of the username when identifying the phenomenon of username abbreviation, which takes a long time and is difficult to apply to large-scale real-time username abbreviation recognition tasks.

Contents of the invention

In view of the above deficiencies, the present invention provides a method for identifying user name abbreviation, which does not need to enumerate user name substrings, reduces the amount of calculation, has low time overhead, and can more efficiently identify user name abbreviation.

In order to solve the problems of the technologies described above, the present invention adopts the following technical solutions:

A method for identifying abbreviated usernames, the steps comprising:

1) Filter the characters in two or more user names, and only keep English letters and numbers;

2) the user name after the above-mentioned filtering is respectively divided into several continuous segments, each selects the first character of each segment to form a new character string;

3) Obtain the length of the longest abbreviation according to the above-mentioned new character string. If the length value is greater than or equal to a given threshold ΔL, it is determined that there is an abbreviation phenomenon among the user names.

Further, the reserved English letters are uniformly converted into lowercase or uppercase.

Further, the segment is a word or a single character.

Further, the segment is obtained by segmenting according to a specified dictionary.

Further, the dictionary includes personal names, place names, object names, invented words or other specified words, and the specified words include nouns, verbs, adjectives, and adverbs.

Further, a dynamic programming algorithm is used to obtain the longest abbreviation length according to the new character string.

Further, the threshold ΔL is the minimum length of the abbreviated form of the user name to be recognized.

Further, when it is desired to recognize the pinyin abbreviation of Chinese personal names, ΔL≥2.

Further, when it is desired to recognize the abbreviated forms of English personal names, ΔL=2.

The beneficial effect of the present invention is that the method provided by the present invention does not need to enumerate all the substrings of the user names, and can automatically identify whether there is abbreviation among the user names. Before identifying and judging the user name, the user name is divided and abbreviated in advance, which reduces the length of the character string itself in the identification process compared with the prior art, thereby reducing the amount of calculation. When identifying user name abbreviations, this method can easily determine by judging the value of the longest abbreviation length, and the acquisition of the longest abbreviation length uses a dynamic programming algorithm. The process of substring reduces a lot of repeated calculations.

Description of drawings

FIG. 1 is a flow chart of a method for identifying abbreviated user names provided in an embodiment.

detailed description

In order to make the above-mentioned features and advantages of the present invention more comprehensible, the following specific embodiments are described in detail in conjunction with the accompanying drawings.

This provides a method for identifying abbreviated usernames. As shown in Figure 1, assuming two usernames a and b are given, it is judged whether there is abbreviated usernames between a and b, including the following steps:

1. Username preprocessing

The user name can usually contain some special characters such as English letters, numbers and underscores. This step is to remove the special characters in the user name. The special characters mentioned in this invention refer to all characters except English letters and numbers, and only English characters are reserved. Letters and numbers, and uniformly convert English letters into lowercase or uppercase. In this embodiment, lowercase is used as an example.

2. Split and abbreviate usernames

According to the dictionary W of words with actual meaning specified by the designator, the user name is divided into several consecutive segments, each segment is a word or a single character in W, and it is necessary to ensure that the number of segmented segments is as small as possible. The designator of the dictionary W is the user, and may be other people. The so-called words with practical meaning, such as the names of people, places, objects or other concepts that have meaning for the designator, such concepts can be nouns, verbs, adjectives, adverbs or other sexual words, or other words created by the designator vocabulary.

The above-mentioned preprocessed user name can be divided by the following method. For the convenience of expression, the given user name is set to u here, and the length is n. Note that S _i represents the i-th character to the n-th character of the user name u The segmentation result of the substring composed of characters, the segmentation steps are as follows:

(1) Initialize i=n+1, S _n+1 ={}.

(2) Let i=i-1, check each word w in the dictionary W in turn, let the length of w be m, if w=u _i u _i+1 ...u _i+m-1 is satisfied, that is, w appears in the user The position of the i-th character in the name u, at this time, if S _i is not assigned or satisfies |S _i+m |+1<|S _i |, then let S _i =w∪S _i+m . If S _i is still unassigned after checking all the words in W, set S _i =u[i]∪S _i+m . Where u[i] represents the i-th character of u.

(3) Repeat step (2) until i=0, at this time S ₀ is the segmentation result of user name u.

After the user name is segmented, the first character of each segment in the segmented result is taken to form a new character string as the abbreviated form of the original user name.

3. Calculate the length of the longest abbreviation

Assuming that the segmentation results of user names a and b obtained after the second step are X _a and X _b respectively, and the abbreviated forms are Y _a and Y _b respectively, it is necessary to obtain the longest abbreviation length m of user names a and b. The length of the longest abbreviation refers to the longest common part of two user name abbreviations that meet a specific condition. The specific condition is that each character of the common part corresponds to the string in the segmentation results of the two users before the abbreviation, all of which satisfy One is a single character and the other is a word. In order to obtain the length m of the longest abbreviation, a dynamic programming algorithm is specially designed, and its formula is as follows:

$\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&le;</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&le;</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; |</span>}{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

Among them, Y _a [i] represents the i-th character of the string Y _a , Y _b [j] represents the j-th character of the string Y _b , |X _a [i]| represents the i-th character in the set X _a The length of the string, |X _b [j]| indicates the length of the jth string in the set X _b , |Y _a | indicates the length of the string Y _a , |Y _b | indicates the length of the string Y _b .

4. Identify username abbreviations

Assuming the given threshold is ΔL, if m≥ΔL, it means that there is abbreviation between user names a and b, otherwise it means that there is no abbreviation between user names a and b.

In combination with the above method, the following two embodiments applied to specific scenarios are given to illustrate the feasibility of this method.

Example 1:

Embodiment 1 is used to identify whether there is abbreviation of pinyin among user names. According to the characteristics of Chinese names, the name must be at least two characters, such as Zhang Wei, Shi Xiaoming, etc. Taking Zhang Wei as an example, the pinyin of his name is ZhangWei or Wei Zhang, and the pinyin abbreviation is random, and From a statistical point of view, the pinyin abbreviation is likely to be the first letter of the name, that is, zw or wz; and the pinyin of Shi Xiaoming's name Shi Xiaoming or Xiaoming Shi, the pinyin abbreviation is likely to be sxm or xms, from the above analysis. It is considered that W is a set of pinyin whose character string length is not less than 2, and ΔL=2. Similarly, if you want to identify the abbreviation of an English name, since English names are generally less used as middle names, that is, English names are at least composed of first name and last name. For example, the English name Sheldon Lee Cooper is often SheldonCooper, and the abbreviation is sc , so the minimum length of the abbreviated form to be recognized can be set to 2, that is, ΔL=2.

Given two user names a=zgxxidian123, b=zhangguoxin012, the present embodiment intends to identify whether there is an abbreviation phenomenon between these two user names by the method provided by the present invention, obtain user name a, The segmentation results of b are respectively X _a = {z, g, x, xi, dian, 1, 2, 3}, X _b = {zhang, guo, xin, 0, 1, 2}, and the abbreviated forms are respectively Y _a =zgxxd123, Y _b =zgx012, further, the length of the longest abbreviation of user names a and b is calculated as m=3 through the third step above. It is known from the above paragraph that ΔL=2, then m≥ΔL, indicating that there is an abbreviation phenomenon between user names a and b.

Example 2:

This embodiment 2 is used to identify whether there is abbreviation of pinyin among user names. From the analysis of the above embodiment, it can be known that W is a set of pinyin whose character string length is not less than 2, and ΔL=2.

Given two user names a=wanxia68 and b=wanter_123, the segmentation results of user names a and b obtained through the method provided in the second step above are respectively X _a ={wan,xia,6,8}, X _b ={ wan, te, r, 1, 2, 3}, the abbreviated forms are respectively Y _a =wx68, Y _b =wtr123, and further, the length m of the longest abbreviation of user name a and b is calculated through the above step 3 =0. Since m<ΔL, it means that there is no abbreviation phenomenon between user names a and b.

The method provided by the invention is to automatically identify whether there is an abbreviation phenomenon in the user name through an algorithm, without enumerating all the substrings of the user name as in the prior art, which is simple and feasible. Before identifying and judging the user name, the user name is divided and abbreviated in advance, which reduces the length of the character string itself in the identification process compared with the prior art, thereby reducing the amount of calculation. When identifying user name abbreviations, this method can easily determine by judging the value of the longest abbreviation length, and the acquisition of the longest abbreviation length uses a dynamic programming algorithm. The process of substring reduces a lot of repeated calculations.

Finally, it should be noted that although the present invention has been disclosed above with embodiments, these embodiments are not intended to limit the present invention, and those skilled in the art can modify or replace it without departing from the spirit and spirit of the present invention. scope, so the protection scope of the present invention shall be determined by the claims.

Claims (9)

1. the method identifying user initials phenomenon, step includes:

1) character in two or more user names is filtered, only retain English alphabet and numeral；

2) user name after above-mentioned filtration is respectively divided into several continuous fragments, respectively chooses the initial character composition one of each fragment New character strings；

3) obtain the longest abbreviation length according to above-mentioned new character strings, if length value is more than or equal to given threshold value Δ L, then judge There is between described user name user initials phenomenon.

Method the most according to claim 1, it is characterised in that be converted into lower case or upper case by unified for the English alphabet of reservation Form.

Method the most according to claim 1, it is characterised in that described fragment is word or single character.

Method the most according to claim 1, it is characterised in that described fragment obtains according to the dictionary segmentation specified.

Method the most according to claim 4, it is characterised in that described dictionary include name, place name, name, fabricate word or Other specify word, and this appointment word includes noun, verb, adjective, adverbial word.

Method the most according to claim 1, it is characterised in that use dynamic programming algorithm to obtain according to described new character strings The longest abbreviation length.

Method the most according to claim 1, it is characterised in that described threshold value Δ L is user initials form to be identified Minimum length.

Method the most according to claim 7, it is characterised in that when Chinese personal name Pinyin abbreviation form to be identified, Δ L >= 2。

Method the most according to claim 7, it is characterised in that when English name-to abbreviated form to be identified, Δ L=2.