WO2017215243A1 - 简历搜索方法及装置 - Google Patents

简历搜索方法及装置 Download PDF

Info

Publication number
WO2017215243A1
WO2017215243A1 PCT/CN2016/113161 CN2016113161W WO2017215243A1 WO 2017215243 A1 WO2017215243 A1 WO 2017215243A1 CN 2016113161 W CN2016113161 W CN 2016113161W WO 2017215243 A1 WO2017215243 A1 WO 2017215243A1
Authority
WO
WIPO (PCT)
Prior art keywords
resume
related word
keyword
search
weight
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2016/113161
Other languages
English (en)
French (fr)
Inventor
李贤�
陈振安
王鹏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Shiyuan Electronics Thecnology Co Ltd
Original Assignee
Guangzhou Shiyuan Electronics Thecnology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangzhou Shiyuan Electronics Thecnology Co Ltd filed Critical Guangzhou Shiyuan Electronics Thecnology Co Ltd
Publication of WO2017215243A1 publication Critical patent/WO2017215243A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2453Query optimisation
    • G06F16/24534Query rewriting; Transformation
    • G06F16/24537Query rewriting; Transformation of operators
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/279Recognition of textual entities
    • G06F40/284Lexical analysis, e.g. tokenisation or collocates

Definitions

  • the invention relates to the field of computer information retrieval, and in particular to a resume search method and device.
  • the frequency of keywords and related words hits during the retrieval process determines the ranking of the searched resumes, and the search results are likely to be biased towards a higher frequency related words.
  • the keyword may be overwhelmed by the related words during the retrieval process, and it is difficult to highlight the role of the keyword as the main search keyword in the process of searching for the resume.
  • the embodiment of the invention provides a resume search method, calculates the weight of keywords and related words in the search process, and improves the accuracy of the search resume.
  • An embodiment of the present invention provides a resume search method, including:
  • the related word set receives a keyword and a related word set for performing a resume search from a resume database; wherein the related word set records a plurality of related words, and a relevance of each related word to the keyword;
  • the weight value of each related word and the weight value of the keyword when performing the resume search are calculated, specifically:
  • Base score formula Calculating a reference score S i of the i-th related word in the related word set; wherein r i is a correlation degree of the i-th related word in the related word set; r min is recorded in the related word set The minimum value of the relevance of all related words; r max is the maximum value of the correlation of all related words;
  • the method further includes:
  • the keyword, the weight value occupied by the keyword, each related word, and the weight value of each of the related words are submitted to a search engine to be used from the resume database Search for a resume, specifically:
  • the keyword and each of the related words included in the resume are displayed in a highlighted form.
  • an embodiment of the present invention further provides a resume search device, including:
  • a receiving module configured to receive keywords and related words used for searching for a resume from a resume database;
  • the related word set records a plurality of related words, and the relevance of each related word to the keyword;
  • a weight calculation module configured to calculate, according to the relevance of each related word and the keyword, a weight value of each related word and a weight value of the keyword when performing a resume search
  • a search module configured to submit, by the keyword, a weight value occupied by the keyword, each of the related words, and a weight value of each of the related words to a search engine, to use the resume database Search for a resume.
  • weight calculation module specifically includes:
  • Benchmark score calculation unit for formulating a score based on a benchmark Calculating a reference score S i of the i-th related word in the related word set; wherein r i is a correlation degree of the i-th related word in the related word set; r min is recorded in the related word set The minimum value of the relevance of all related words; r max is the maximum value of the correlation among all related words;
  • a sum calculation unit for summing the reference scores of all the related words to obtain a total score of the reference scores sum
  • a related word weight calculation unit configured to determine whether a reference score of the i-th related word is greater than W A /n; if yes, according to a second weight formula Calculating a weight value W i of the i-th related word when performing a resume search; if not, according to the third weight formula Calculating a weight value W i of the i-th related word when performing a resume search; wherein n is a weight coefficient; k is the number of related words in the related word set.
  • the weight calculation module further includes:
  • a mean value calculation unit configured to: according to the mean value formula, before determining whether the reference score of the i-th related word is greater than W A /n Calculate the mean of the benchmark score among them, The average of the correlations of all related words;
  • a determining unit configured to determine, for the i-th related word in the related word set, whether the reference score S i of the i-th related word is greater than the reference score mean
  • An update unit configured to: when determining that the reference score S i of the ith related word is greater than the reference score mean By updating the formula Updating the reference score S i of the i-th related word.
  • searching module specifically includes:
  • a weight associating unit configured to associate, according to a weight association format of the search engine, a weight value occupied by the keyword and the keyword as a first combination, and a weight value occupied by each of the related words Corresponding related words are associated with the second combination;
  • Searching a display unit configured to submit the first combination and the second combination to a search engine, for the search engine to search for a resume from the resume database, and display according to a sorting algorithm built in the search engine The searched resume; wherein the keyword and each of the related words included in the resume are displayed in a highlighted form.
  • the resume search method and device provided by the embodiment of the present invention divides the weight value occupied by the keyword and the related word in the resume search process, so that the resume can be retrieved through the related word, and the keyword can be generalized and highlighted.
  • the main status of the keyword in the retrieval process so that the resume matching the domain corresponding to the keyword can be found more accurately in the retrieval process, making the retrieved resume more reasonable.
  • FIG. 1 is a schematic flow chart of an embodiment of a resume search method provided by the present invention
  • FIG. 2 is a schematic structural diagram of an embodiment of a resume search device provided by the present invention.
  • FIG. 3 is a schematic structural diagram of an embodiment of a weight calculation module of a resume search device provided by the present invention.
  • FIG. 4 is a schematic structural diagram of an embodiment of a search module of a resume search device provided by the present invention.
  • FIG. 1 is a schematic flowchart diagram of an embodiment of a resume search method provided by the present invention, where the method includes steps S1 to S3, as follows:
  • S1 Receive a keyword and a related word set for performing a resume search from a resume database; wherein the related word set records a plurality of related words, and a relevance of each related word to the keyword;
  • step S2 the implementation manner of the foregoing step S2 is as follows:
  • Base score formula Calculating a reference score S i of the i-th related word in the related word set; wherein r i is a correlation degree of the i-th related word in the related word set; r min is recorded in the related word set The minimum value of the relevance of all related words; r max is the maximum value of the correlation among all related words;
  • the purpose of calculating the benchmark score of each related word in the related word set is to base the correlation degree of the related words on the same benchmark, and represent the degree of correlation between the related words and the keywords in the form of a score.
  • the reference value is preferably r max -r min +1; the first weight formula adopts a logarithmic formula, and when the total value of the benchmark score grows too fast, the weight value of the keyword still maintains a moderate growth, and thus, the relevant words can be In a relatively large number of cases, avoid the occurrence of excessive weight of keywords; when calculating the weight of related words, the second weight formula and the third weight formula are used respectively to calculate, considering the relevant words.
  • the weight value of the related word is greater than the weight value of the keyword, or is much smaller than the weight value of the keyword, that is, during the process of searching for the resume, It can avoid the situation that the relevant words are overwhelmed by keywords and the keywords are too prominent; the setting of the weight coefficient is derived from how many lexicons the related words are related to. Is determined, e.g., acquired from a related word dictionary, the value of n is 2; related words are acquired from two thesaurus, the value of n is 4.
  • the method further includes:
  • the correlation degree of the related words in the related word set is based on the same benchmark, and the benchmark is the same as the above benchmark, and the purpose is to represent the average correlation between the related word set and the keyword in the form of a score.
  • Degree when the degree of correlation between a related word and a keyword (ie, the benchmark score) is greater than the average degree of correlation between the related word set and the keyword (ie, the mean value of the benchmark score), the related word may be related to the keyword Degree, the difference between the correlation degree and the average degree of relevance of the keyword, the degree of difference between the two degrees of relevance is magnified 10 times as the score of the degree of relevance of the related word to the keyword; otherwise, the correlation between the related word and the keyword is maintained The original score of the degree, which more accurately expresses the relevance of the related words to the keywords.
  • step S3 is:
  • the keyword and each of the related words included in the resume are displayed in a highlighted form.
  • the commonly used search engine is the solr search engine, and the above weight association format can be associated according to the following format: keyword ⁇ keyword weight, related word 1 ⁇ related word 1 weight, related word 2 ⁇ related words 2 weights...
  • the resume search method of the embodiment of the invention is implemented, and the relationship between the related words and the keywords is made more reasonable by adjusting the correlation degree between the related words and the keywords, and the keywords and the reference points are calculated according to the reference score indicating the degree of relevance.
  • the weight value of the related words in the resume search process so that the resume can be retrieved through the related words, and the keywords can be generalized, and the subject position of the keywords in the retrieval process can be highlighted, so that the search process can be more Accurately find the resume that matches the field corresponding to the keyword, making the retrieved resume more reasonable.
  • the resume search device includes:
  • the receiving module 10 is configured to receive a keyword and a related word set for performing a resume search from a resume database, wherein the related word set records a plurality of related words, and each related word is related to the keyword degree;
  • the weight calculation module 20 is configured to calculate, according to the relevance of each related word and the keyword, a weight value of each related word and a weight value occupied by the keyword when performing a resume search;
  • the search module 30 is configured to submit the keyword, the weight value occupied by the keyword, each of the related words, and the weight value of each of the related words to the search engine, to Search for a resume in the resume database.
  • FIG. 3 is a schematic structural diagram of an embodiment of a weight calculation module of a resume search device provided by the present invention.
  • the weight calculation module 20 specifically includes:
  • a reference score calculation unit 21 for using a reference score formula Calculating a reference score S i of the i-th related word in the related word set; wherein r i is a correlation degree of the i-th related word in the related word set; r min is recorded in the related word set The minimum value of the relevance of all related words; r max is the maximum value of the correlation among all related words;
  • the sum calculation unit 22 is configured to sum the reference scores of all the related words to obtain a reference score total value sum
  • the related word weight calculation unit 24 is configured to determine whether the reference score of the i-th related word is greater than W A /n; if yes, according to the second weight formula Calculating a weight value W i of the i-th related word when performing a resume search; if not, according to the third weight formula Calculating a weight value W i of the i-th related word when performing a resume search; wherein n is a weight coefficient; k is the number of related words in the related word set.
  • the weight calculation module further includes:
  • the mean value calculating unit 25 is configured to: according to the mean value formula, before determining whether the reference score of the ith related word is greater than W A /n Calculate the mean of the benchmark score among them, The average of the correlations of all related words;
  • the determining unit 26 is configured to determine, for the i-th related word in the related word set, whether the reference score S i of the i-th related word is greater than the reference score mean
  • the updating unit 27 is configured to: when determining that the reference score S i of the ith related word is greater than the reference score mean By updating the formula Updating the reference score S i of the i-th related word.
  • FIG. 4 is a schematic structural diagram of an embodiment of a search module of a resume search device provided by the present invention; the search module 30 specifically includes:
  • the weight associating unit 31 is configured to associate, according to a weight association format of the search engine, a weight value occupied by the keyword and the keyword as a first combination, and a weight of each of the related words The value is associated with its corresponding related word as a second combination;
  • a search display unit 32 configured to submit the first combination and the second combination to a search engine, for the search engine to search for a resume from the resume database, and according to a sorting algorithm built in the search engine
  • the searched resume is displayed; wherein the keyword and each of the related words included in the resume are displayed in a highlighted form.
  • the resume searching device of the embodiment of the present invention optimizes the relationship between the related words and the keywords by adjusting the degree of correlation between the related words and the keywords, and calculates keywords and based on the reference score indicating the degree of relevance.
  • the weight value of the related words in the resume search process so that the resume can be retrieved through the related words, and the keywords can be generalized, and the subject position of the keywords in the retrieval process can be highlighted, so that the search process can be more Accurately find the resume that matches the field corresponding to the keyword, making the retrieved resume more reasonable.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • General Health & Medical Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

一种简历搜索方法和装置,计算出关键词和相关词在搜索过程的权重,提高检索简历的准确度;该方法包括:接收用于从简历数据库中进行简历搜索的关键词和相关词集(S1);其中,所述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值(S2);将所述关键词、所述关键词所占的权重值、每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历(S3)。

Description

简历搜索方法及装置 技术领域
本发明涉及计算机信息检索领域,尤其涉及一种简历搜索方法及装置。
背景技术
在以往对简历的搜索匹配过程中,存在两种方法:一种是直接关键词检索;另一种是关键词+相关词检索;但是发明人在实施过程发现上述方案存在以下问题:
对于第一种方案,仅考虑关键词的检索过程中出现的词频来获取检索结果,即,偏向与该关键词相关的技能的简历排序在前,难以综合考虑简历的综合技能,不便于综合评估简历本人的综合技能。
对于第二种方案,在检索过程中关键词和相关词命中的频次决定了搜索出来的简历的排序,则检索结果极有可能偏向某个频次较高的相关词,另一方面,当相关词的数量大大多于关键词时,则在检索过程中关键词可能会被相关词淹没,难以突出关键词相对于相关词在搜索出简历过程中作为主要作检索关键词的作用。
发明内容
本发明实施例提出一种简历搜索方法,计算出关键词和相关词在搜索过程的权重,提高检索简历的准确度。
本发明实施例提供一种简历搜索方法,包括:
接收用于从简历数据库中进行简历搜索的关键词和相关词集;其中,所述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;
根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值;
将所述关键词、所述关键词所占的权重值、所述每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历。
进一步地,所述根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值,具体为:
根据基准分值公式
Figure PCTCN2016113161-appb-000001
计算出所述相关词集中第i个相关词的基准分值Si;其中,ri为所述相关词集中的第i个相关词的相关度;rmin为所述相关词集所记载的所有相关词的相关度的最小值;rmax为所述所有相关词的相关度的最大值;
对所述所有相关词的基准分值进行求和,获得基准分值总值sum;
根据第一权重公式WA=5+log1.5(sum+1),计算出所述关键词在进行简历搜索时所占的权重值WA;其中,A为所述关键词;
判断所述第i个相关词的基准分值是否大于WA/n;若是,则根据第二权重公式
Figure PCTCN2016113161-appb-000002
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;若否,则根据第三权重公式
Figure PCTCN2016113161-appb-000003
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;其中,n为权重系数,k为所述相关词集中的相关词的数量。
再进一步地,在所述判断所述第i个相关词的基准分值是否大于WA/n之前,还包括:
根据均值公式
Figure PCTCN2016113161-appb-000004
计算基准分值均值
Figure PCTCN2016113161-appb-000005
其中,
Figure PCTCN2016113161-appb-000006
为所述所有相关词的相关度的平均值;
对于所述相关词集中的第i个相关词,判断所述第i个相关词的基准分值Si是否大于所述基准分值均值
Figure PCTCN2016113161-appb-000007
若是,通过更新公式
Figure PCTCN2016113161-appb-000008
更新所述第i个相关词的基准分值Si
更进一步地,所述将所述关键词、所述关键词所占的的权重值、每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历,具体为:
根据搜索引擎的权重关联格式,将所述关键词所占的权重值和所述关键词相关联作为第一组合,以及将所述每一个相关词所占的权重值与其对应的相关词相关联作为第二组合;
将所述第一组合和所述第二组合提交给搜索引擎,以供所述搜索引擎从所述简历数据库中搜索出简历,并根据所述搜索引擎内置的排序算法显示搜索出的简历;其中,所述简历中包含的所述关键词和所述每一个相关词以高亮形式显示。
相应地,本发明实施例还提供一种简历搜索装置,包括:
接收模块,用于接收用于从简历数据库中进行简历搜索的关键词和相关词集;其中,所 述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;
权重计算模块,用于根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值;
搜索模块,用于将所述关键词、所述关键词所占的权重值、所述每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历。
进一步地,所述权重计算模块具体包括:
基准分值计算单元,用于根据基准分值公式
Figure PCTCN2016113161-appb-000009
计算出所述相关词集中第i个相关词的基准分值Si;其中,ri为所述相关词集中的第i个相关词的相关度;rmin为在所述相关词集中所记载的所有相关词的相关度的最小值;rmax为在所述所有相关词的相关度的最大值;
求和计算单元,用于对所述所有相关词的基准分值进行求和,获得基准分值总值sum;
关键词权重计算单元,用于根据第一权重公式WA=5+log1.5(sum+1),计算出所述关键词在进行简历搜索时所占的权重值WA;其中,A为所述关键词;
相关词权重计算单元,用于判断所述第i个相关词的基准分值是否大于WA/n;若是,则根据第二权重公式
Figure PCTCN2016113161-appb-000010
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;若否,则根据第三权重公式
Figure PCTCN2016113161-appb-000011
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;其中,n为权重系数;k为所述相关词集中的相关词的数量。
再进一步地,所述权重计算模块还包括:
均值计算单元,用于在所述判断所述第i个相关词的基准分值是否大于WA/n之前,根据均值公式
Figure PCTCN2016113161-appb-000012
计算基准分值均值
Figure PCTCN2016113161-appb-000013
其中,
Figure PCTCN2016113161-appb-000014
为所述所有相关词的相关度的平均值;
判断单元,用于对于所述相关词集中的第i个相关词,判断所述第i个相关词的基准分值Si是否大于所述基准分值均值
Figure PCTCN2016113161-appb-000015
更新单元,用于当判断所述第i个相关词的基准分值Si大于所述基准分值均值
Figure PCTCN2016113161-appb-000016
时,通过更新公式
Figure PCTCN2016113161-appb-000017
更新所述第i个相关词的基准分值Si
更进一步地,所述搜索模块具体包括:
权重相联单元,用于根据搜索引擎的权重关联格式,将所述关键词所占的权重值和所述关键词相关联作为第一组合,以及将所述每一个相关词所占的权重值与其对应的相关词相关联作为第二组合;
搜索显示单元,用于将所述第一组合和所述第二组合提交给搜索引擎,以供所述搜索引擎从所述简历数据库中搜索出简历,并根据所述搜索引擎内置的排序算法显示搜索出的简历;其中,所述简历中包含的所述关键词和所述每一个相关词以高亮形式显示。
实施本发明实施例,具有如下有益效果:
本发明实施例提供的简历搜索方法和装置,通过划分关键词和相关词在简历搜索过程中所占的权重值,使得既能通过相关词检索到的简历,同时也能泛化关键词,突出关键词在检索过程中的主体地位,从而在检索过程中可以更准确地找到与关键词对应领域下匹配的简历,使得检索出来的简历更为合理。
附图说明
图1是本发明提供的简历搜索方法的一个实施例的流程示意图;
图2是本发明提供的简历搜索装置的一个实施例的结构示意图;
图3是本发明提供的简历搜索装置的权重计算模块的一个实施例的结构示意图;
图4是本发明提供的简历搜索装置的搜索模块的一个实施例的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
参见图1,是本发明提供的简历搜索方法的一个实施例的流程示意图,该方法包括步骤S1至S3,具体如下:
S1,接收用于从简历数据库中进行简历搜索的关键词和相关词集;其中,所述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;
S2,根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值;
S3,将所述关键词、所述关键词所占的权重值、所述每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历。
在具体实施时,上述步骤S2的实施方式如下:
根据基准分值公式
Figure PCTCN2016113161-appb-000018
计算出所述相关词集中第i个相关词的基准分值Si;其中,ri为所述相关词集中的第i个相关词的相关度;rmin为在所述相关词集中所记载的所有相关词的相关度的最小值;rmax为在所述所有相关词的相关度的最大值;
对所述所有相关词的基准分值进行求和,获得基准分值总值sum;
根据第一权重公式WA=5+log1.5(sum+1),计算出所述关键词在进行简历搜索时所占的权重值WA;其中,A为所述关键词;
判断所述第i个相关词的基准分值是否大于WA/n;若是,则根据第二权重公式
Figure PCTCN2016113161-appb-000019
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;若否,则根据第三权重公式
Figure PCTCN2016113161-appb-000020
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;其中,n为权重系数,k为所述相关词集中的相关词的数量。
需要说明的是,计算相关词集中每个相关词的基准分值的目的是将相关词的相关度基于同一个基准,以一个分值的形式,表示该相关词与关键词的相关程度,该基准值优选为rmax-rmin+1;第一权重公式采用对数形式的公式,在基准分值总值增长过快时,关键词的权重值仍保持缓和增长,因而,可以在相关词相对较多的情况下,避免关键词所占权重过大这种情况的出现;在进行相关词的权重计算时,分别采用第二权重公式和第三权重公式进行计算,是考虑到当相关词的基准分值过大和过小这两种情况出现时,避免计算出该相关词的权重值大于关键词的权重值,或者是大大小于关键词的权重值,也就是说在检索简历过程中,能避免相关词过多把关键词淹没、以及出现关键词过于突出的情况;权重系数的设置,来源于该相关词是从多少个词库中获取相关词的而决定,例如,从一个词库中获取相关词,则n的取值为2;分别从两个词库中获取相关词,则n的取值为4。
再进一步地,在上述步骤S2的实施方式中的在所述判断所述第i个相关词的基准分值是否大于WA/n之前,还包括:
根据均值公式
Figure PCTCN2016113161-appb-000021
计算基准分值均值
Figure PCTCN2016113161-appb-000022
其中,
Figure PCTCN2016113161-appb-000023
为所述所有相关词的相关度的平均值;
对于所述相关词集中的第i个相关词,判断所述第i个相关词的基准分值Si是否大于所述 基准分值均值
Figure PCTCN2016113161-appb-000024
若是,通过更新公式
Figure PCTCN2016113161-appb-000025
更新所述第i个相关词的基准分值Si
需要说明的是,将该相关词集中的相关词的相关度均值基于同一个基准,且该基准与上述基准相同,目的是以一个分值的形式,表示该相关词集与关键词的平均相关程度;当一个相关词与关键词的相关程度(即基准分值)大于该相关词集与关键词的平均相关程度(即,基准分值均值)时,可将该相关词与关键词的相关程度,与该相关词集与关键词的平均相关程度两个相关程度的差异值放大10倍作为该相关词与关键词的相关程度的分值;反之,则保持该相关词与关键词的相关程度的原分值,这种方式更能确切地表达该相关词与关键词的相关程度。
更进一步地,上述步骤S3的具体实施方式为:
根据搜索引擎的权重关联格式,将所述关键词所占的权重值和所述关键词相关联作为第一组合,以及将所述每一个相关词所占的权重值与其对应的相关词相关联作为第二组合;
将所述第一组合和所述第二组合提交给搜索引擎,以供所述搜索引擎从所述简历数据库中搜索出简历,并根据所述搜索引擎内置的排序算法显示搜索出的简历;其中,所述简历中包含的所述关键词和所述每一个相关词以高亮形式显示。
需要说明的是,一般采用的搜索引擎为solr搜索引擎,则上述权重关联格式,可依据如下格式进行关联:关键词^关键词权重、相关词1^相关词1权重、相关词2^相关词2权重......
实施本发明实施例的简历搜索方法,通过调整相关词与关键词的相关程度,使相关词与关键词之间的关系更为合理,并依据表示该相关程度的基准分值计算出关键词和相关词在简历搜索过程中所占的权重值,使得既能通过相关词检索到的简历,同时也能泛化关键词,突出关键词在检索过程中的主体地位,从而在检索过程中可以更准确地找到与关键词对应领域下匹配的简历,使得检索出来的简历更为合理。
相应地,参见图2,是本发明提供的简历搜索装置的一个实施例的结构示意图,该简历搜索装置,包括:
接收模块10,用于接收用于从简历数据库中进行简历搜索的关键词和相关词集;其中,所述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;
权重计算模块20,用于根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值;
搜索模块30,用于将所述关键词、所述关键词所占的的权重值、所述每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历。
进一步地,如图3所示,图3是本发明提供的简历搜索装置的权重计算模块的一个实施例的结构示意图,所述权重计算模块20具体包括:
基准分值计算单元21,用于根据基准分值公式
Figure PCTCN2016113161-appb-000026
计算出所述相关词集中第i个相关词的基准分值Si;其中,ri为所述相关词集中的第i个相关词的相关度;rmin为在所述相关词集中所记载的所有相关词的相关度的最小值;rmax为在所述所有相关词的相关度的最大值;
求和计算单元22,用于对所述所有相关词的基准分值进行求和,获得基准分值总值sum;
关键词权重计算单元23,用于根据第一权重公式WA=5+log1.5(sum+1),计算出所述关键词在进行简历搜索时所占的权重值WA;其中,A为所述关键词;
相关词权重计算单元24,用于判断所述第i个相关词的基准分值是否大于WA/n;若是,则根据第二权重公式
Figure PCTCN2016113161-appb-000027
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;若否,则根据第三权重公式
Figure PCTCN2016113161-appb-000028
计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;其中,n为权重系数;k为所述相关词集中的相关词的数量。
再进一步地,所述权重计算模块还包括:
均值计算单元25,用于在所述判断所述第i个相关词的基准分值是否大于WA/n之前,根据均值公式
Figure PCTCN2016113161-appb-000029
计算基准分值均值
Figure PCTCN2016113161-appb-000030
其中,
Figure PCTCN2016113161-appb-000031
为所述所有相关词的相关度的平均值;
判断单元26,用于对于所述相关词集中的第i个相关词,判断所述第i个相关词的基准分值Si是否大于所述基准分值均值
Figure PCTCN2016113161-appb-000032
更新单元27,用于当判断所述第i个相关词的基准分值Si是大于所述基准分值均值
Figure PCTCN2016113161-appb-000033
时,通过更新公式
Figure PCTCN2016113161-appb-000034
更新所述第i个相关词的基准分值Si
更进一步地,如图4所示,图4是本发明提供的简历搜索装置的搜索模块的一个实施例的结构示意图;所述搜索模块30具体包括:
权重相联单元31,用于根据搜索引擎的权重关联格式,将所述关键词所占的权重值和所述关键词相关联作为第一组合,以及将所述每一个相关词所占的权重值与其对应的相关词相关联作为第二组合;
搜索显示单元32,用于将所述第一组合和所述第二组合提交给搜索引擎,以供所述搜索引擎从所述简历数据库中搜索出简历,并根据所述搜索引擎内置的排序算法显示搜索出的简历;其中,所述简历中包含的所述关键词和所述每一个相关词以高亮形式显示。
实施本发明实施例的简历搜索装置,通过调整相关词与关键词的相关程度,使相关词与关键词之间的关系更为合理,并依据表示该相关程度的基准分值计算出关键词和相关词在简历搜索过程中所占的权重值,使得既能通过相关词检索到的简历,同时也能泛化关键词,突出关键词在检索过程中的主体地位,从而在检索过程中可以更准确地找到与关键词对应领域下匹配的简历,使得检索出来的简历更为合理。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成,所述的程序可存储于一计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory,ROM)或随机存储记忆体(Random Access Memory,RAM)等。
以上所述是本发明的优选实施方式,应当指出,对于本技术领域的普通技术人员来说,在不脱离本发明原理的前提下,还可以做出若干改进和润饰,这些改进和润饰也视为本发明的保护范围。

Claims (8)

  1. 一种简历搜索方法,其特征在于,包括:
    接收用于从简历数据库中进行简历搜索的关键词和相关词集;其中,所述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;
    根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值;
    将所述关键词、所述关键词所占的权重值、所述每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历。
  2. 如权利要求1所述简历搜索方法,其特征在于,所述根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值,具体为:
    根据基准分值公式
    Figure PCTCN2016113161-appb-100001
    计算出所述相关词集中第i个相关词的基准分值Si;其中,ri为所述相关词集中的第i个相关词的相关度;rmin为所述相关词集所记载的所有相关词的相关度的最小值;rmax为所述所有相关词的相关度的最大值;
    对所述所有相关词的基准分值进行求和,获得基准分值总值sum;
    根据第一权重公式WA=5+log1.5(sum+1),计算出所述关键词在进行简历搜索时所占的权重值WA;其中,A为所述关键词;
    判断所述第i个相关词的基准分值是否大于WA/n;若是,则根据第二权重公式
    Figure PCTCN2016113161-appb-100002
    计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;若否,则根据第三权重公式
    Figure PCTCN2016113161-appb-100003
    计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;其中,n为权重系数,k为所述相关词集中的相关词的数量。
  3. 如权利要求2所述的简历搜索方法,其特征在于,在所述判断所述第i个相关词的基准分值是否大于WA/n之前,还包括:
    根据均值公式
    Figure PCTCN2016113161-appb-100004
    计算基准分值均值
    Figure PCTCN2016113161-appb-100005
    其中,
    Figure PCTCN2016113161-appb-100006
    为所述所有相关词的相关度的平均值;
    对于所述相关词集中的第i个相关词,判断所述第i个相关词的基准分值Si是否大于所述基准分值均值
    Figure PCTCN2016113161-appb-100007
    若是,通过更新公式
    Figure PCTCN2016113161-appb-100008
    更新所述第i个相关词的基准分值Si
  4. 如权利要求1所述的简历搜索方法,其特征在于,所述将所述关键词、所述关键词所占的权重值、每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历,具体为:
    根据搜索引擎的权重关联格式,将所述关键词所占的权重值和所述关键词相关联作为第一组合,以及将所述每一个相关词所占的权重值与其对应的相关词相关联作为第二组合;
    将所述第一组合和所述第二组合提交给搜索引擎,以供所述搜索引擎从所述简历数据库中搜索出简历,并根据所述搜索引擎内置的排序算法显示搜索出的简历;其中,所述简历中包含的所述关键词和所述每一个相关词以高亮形式显示。
  5. 一种简历搜索装置,其特征在于,包括:
    接收模块,用于接收用于从简历数据库中进行简历搜索的关键词和相关词集;其中,所述相关词集记载有多个相关词,以及每一个相关词与所述关键词的相关度;
    权重计算模块,用于根据每一个相关词与所述关键词的相关度,计算出在进行简历搜索时每一个相关词所占的权重值及所述关键词所占的权重值;
    搜索模块,用于将所述关键词、所述关键词所占的权重值、所述每一个相关词和所述每一个相关词所占的权重值提交给搜索引擎,以从所述简历数据库中搜索出简历。
  6. 如权利要求5所述简历搜索装置,其特征在于,所述权重计算模块具体包括:
    基准分值计算单元,用于根据基准分值公式
    Figure PCTCN2016113161-appb-100009
    计算出所述相关词集中第i个相关词的基准分值Si;其中,ri为所述相关词集中的第i个相关词的相关度;rmin为在所述相关词集中所记载的所有相关词的相关度的最小值;rmax为在所述所有相关词的相关度的最大值;
    求和计算单元,用于对所述所有相关词的基准分值进行求和,获得基准分值总值sum;
    关键词权重计算单元,用于根据第一权重公式WA=5+log1.5(sum+1),计算出所述关键词在进行简历搜索时所占的权重值WA;其中,A为所述关键词;
    相关词权重计算单元,用于判断所述第i个相关词的基准分值是否大于WA/n;若是,则根据第二权重公式
    Figure PCTCN2016113161-appb-100010
    计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;若否,则根据第三权重公式
    Figure PCTCN2016113161-appb-100011
    计算出所述第i个相关词在进行简历搜索时所占的权重值Wi;其中,n为权重系数;k为所述相关词集中的相关词的数量。
  7. 如权利要求6所述的简历搜索装置,其特征在于,所述权重计算模块还包括:
    均值计算单元,用于在所述判断所述第i个相关词的基准分值是否大于WA/n之前,根据均值公式
    Figure PCTCN2016113161-appb-100012
    计算基准分值均值
    Figure PCTCN2016113161-appb-100013
    其中,
    Figure PCTCN2016113161-appb-100014
    为所述所有相关词的相关度的平均值;
    判断单元,用于对于所述相关词集中的第i个相关词,判断所述第i个相关词的基准分值Si是否大于所述基准分值均值
    Figure PCTCN2016113161-appb-100015
    更新单元,用于当判断所述第i个相关词的基准分值Si大于所述基准分值均值
    Figure PCTCN2016113161-appb-100016
    时,通过更新公式
    Figure PCTCN2016113161-appb-100017
    更新所述第i个相关词的基准分值Si
  8. 如权利要求5所述的简历搜索装置,其特征在于,所述搜索模块具体包括:
    权重相联单元,用于根据搜索引擎的权重关联格式,将所述关键词所占的权重值和所述关键词相关联作为第一组合,以及将所述每一个相关词所占的权重值与其对应的相关词相关联作为第二组合;
    搜索显示单元,用于将所述第一组合和所述第二组合提交给搜索引擎,以供所述搜索引擎从所述简历数据库中搜索出简历,并根据所述搜索引擎内置的排序算法显示搜索出的简历;其中,所述简历中包含的所述关键词和所述每一个相关词以高亮形式显示。
PCT/CN2016/113161 2016-06-17 2016-12-29 简历搜索方法及装置 Ceased WO2017215243A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610445490.5A CN106095982B (zh) 2016-06-17 2016-06-17 简历搜索方法及装置
CN201610445490.5 2016-06-17

Publications (1)

Publication Number Publication Date
WO2017215243A1 true WO2017215243A1 (zh) 2017-12-21

Family

ID=57237572

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/113161 Ceased WO2017215243A1 (zh) 2016-06-17 2016-12-29 简历搜索方法及装置

Country Status (2)

Country Link
CN (1) CN106095982B (zh)
WO (1) WO2017215243A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106095982B (zh) * 2016-06-17 2019-03-29 广州视源电子科技股份有限公司 简历搜索方法及装置
CN105956195B (zh) * 2016-06-17 2019-03-29 广州视源电子科技股份有限公司 简历搜索方法和装置
CN107357917B (zh) * 2017-07-20 2020-04-07 北京拉勾科技有限公司 一种简历搜索方法及计算设备
CN108021641B (zh) * 2017-11-29 2019-07-19 有米科技股份有限公司 应用的关联关键词拓展的方法和装置
CN110084571A (zh) * 2019-05-08 2019-08-02 软通智慧科技有限公司 一种简历筛选方法、装置、服务器和介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007034874A (ja) * 2005-07-29 2007-02-08 Nomura Research Institute Ltd 検索システム及び検索方法
CN103425687A (zh) * 2012-05-21 2013-12-04 阿里巴巴集团控股有限公司 一种基于关键词的检索方法和系统
CN104216931A (zh) * 2013-05-29 2014-12-17 酷盛(天津)科技有限公司 实时推荐系统及方法
CN105956195A (zh) * 2016-06-17 2016-09-21 广州视源电子科技股份有限公司 简历搜索方法和装置
CN106095982A (zh) * 2016-06-17 2016-11-09 广州视源电子科技股份有限公司 简历搜索方法及装置
CN106126589A (zh) * 2016-06-17 2016-11-16 广州视源电子科技股份有限公司 简历搜索方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007034874A (ja) * 2005-07-29 2007-02-08 Nomura Research Institute Ltd 検索システム及び検索方法
CN103425687A (zh) * 2012-05-21 2013-12-04 阿里巴巴集团控股有限公司 一种基于关键词的检索方法和系统
CN104216931A (zh) * 2013-05-29 2014-12-17 酷盛(天津)科技有限公司 实时推荐系统及方法
CN105956195A (zh) * 2016-06-17 2016-09-21 广州视源电子科技股份有限公司 简历搜索方法和装置
CN106095982A (zh) * 2016-06-17 2016-11-09 广州视源电子科技股份有限公司 简历搜索方法及装置
CN106126589A (zh) * 2016-06-17 2016-11-16 广州视源电子科技股份有限公司 简历搜索方法及装置

Also Published As

Publication number Publication date
CN106095982A (zh) 2016-11-09
CN106095982B (zh) 2019-03-29

Similar Documents

Publication Publication Date Title
US9141728B2 (en) Methods and systems for creating and using an adaptive thesaurus
US8122043B2 (en) System and method for using an exemplar document to retrieve relevant documents from an inverted index of a large corpus
WO2017215243A1 (zh) 简历搜索方法及装置
US8352474B2 (en) System and method for retrieving information using a query based index
WO2017215242A1 (zh) 简历搜索方法及装置
US20110184893A1 (en) Annotating queries over structured data
JP6299596B2 (ja) クエリ類似度評価システム、評価方法、及びプログラム
GB2581464A (en) Supporting evidence retrieval for complex answers
JP2013522720A (ja) 単語情報エントロピの決定
WO2021082123A1 (zh) 信息推荐方法及装置、电子设备
WO2019218527A1 (zh) 多系统相结合的自然语言处理方法及装置
CN105956195B (zh) 简历搜索方法和装置
CN111125299B (zh) 一种基于用户行为分析的动态词库更新方法
CN106570196B (zh) 视频节目的搜索方法和装置
WO2018090468A1 (zh) 视频节目的搜索方法和装置
CN111737413A (zh) 基于概念网语义的反馈模型信息检索方法、系统及介质
CN115203379A (zh) 检索方法、装置、计算机设备、存储介质和程序产品
CN108733745A (zh) 一种基于医学知识的查询扩展方法
JP5367632B2 (ja) 知識量推定装置及びプログラム
US10984005B2 (en) Database search apparatus and method of searching databases
CN110362813B (zh) 基于bm25的搜索相关性度量方法、存储介质、设备及系统
CN113590736B (zh) 索引管理方法、装置、电子设备和可读存储介质
TW201516718A (zh) 調整大量資料檢索結果的方法
JP7305077B2 (ja) 情報処理装置、要約文出力方法、及び要約文出力プログラム
US20200117735A1 (en) Method for identifying complex textual patterns containing keywords within data records

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16905350

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16905350

Country of ref document: EP

Kind code of ref document: A1