CN112612961B - Information searching method, device, storage medium and computer equipment - Google Patents

Information searching method, device, storage medium and computer equipment Download PDF

Info

Publication number
CN112612961B
CN112612961B CN202011576921.4A CN202011576921A CN112612961B CN 112612961 B CN112612961 B CN 112612961B CN 202011576921 A CN202011576921 A CN 202011576921A CN 112612961 B CN112612961 B CN 112612961B
Authority
CN
China
Prior art keywords
score
company
matching
entity
search
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.)
Active
Application number
CN202011576921.4A
Other languages
Chinese (zh)
Other versions
CN112612961A (en
Inventor
张伟望
刘炎
覃建策
陈邦忠
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.)
Perfect World Beijing Software Technology Development Co Ltd
Original Assignee
Perfect World Beijing Software Technology Development 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 Perfect World Beijing Software Technology Development Co Ltd filed Critical Perfect World Beijing Software Technology Development Co Ltd
Priority to CN202011576921.4A priority Critical patent/CN112612961B/en
Publication of CN112612961A publication Critical patent/CN112612961A/en
Application granted granted Critical
Publication of CN112612961B publication Critical patent/CN112612961B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9532—Query formulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • G06Q10/105—Human resources
    • G06Q10/1053—Employment or hiring

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Operations Research (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Mathematical Physics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses an information searching method, an information searching device, a storage medium and computer equipment, which relate to the technical field of information and mainly can utilize entities contained in search sentences to accurately match different company positions so as to ensure that the company positions meeting the requirements of users can be accurately provided for users. The method comprises the following steps: acquiring a search sentence input by a user; identifying a first type entity and a second type entity contained in the search statement; obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; and displaying the company positions meeting the preset top ranking conditions according to the similarity total score. The method is mainly suitable for searching company positions.

Description

Information searching method, device, storage medium and computer equipment
Technical Field
The present invention relates to the field of information technologies, and in particular, to an information searching method, an information searching device, a storage medium, and a computer device.
Background
With the development of internet technology, network application and network recruitment have become the main way for job seekers to recruit employees by using personnel units, and the job seekers seek the job positions meeting the requirements of the job seekers by inputting corresponding search sentences in job seekers websites.
Currently, in the job searching process, search sentences input by a user are generally directly matched with different company job positions so as to screen out the job positions meeting the requirements of job seekers and provide the job positions for the job seekers. However, since the search term input by the user is usually a spoken or unstructured search term, the spoken or unstructured search term cannot accurately reflect the actual intention that the user wants to express, and the spoken or unstructured search term is unfavorable for accurate matching with different company positions, and the matching accuracy is low, so that the company positions meeting the requirements of the user cannot be accurately provided for the user.
Disclosure of Invention
The invention provides a job position searching method, a job position searching device, a storage medium and computer equipment, which mainly aim at accurately matching entities contained in search sentences with different company job positions by carrying out entity identification on the spoken or unstructured search sentences input by a user so as to ensure that the company job positions meeting the requirements of the user are accurately provided for the user.
According to a first aspect of the present invention, there is provided an information search method comprising:
acquiring a search sentence input by a user;
identifying a first type entity and a second type entity contained in the search statement;
obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list;
adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position;
and displaying the company positions meeting the preset top ranking conditions according to the similarity total score.
Optionally, after the calculating the similarity scores of the first type of entity and the second type of entity and the field information related to the company positions in the search list, the method further includes:
according to the category information corresponding to the field information, a preset anti-cheating adjustment algorithm is adopted to respectively adjust similarity scores of the first type entity and the second type entity and the field information, and adjusted similarity scores corresponding to the first type entity and the second type entity are obtained;
The step of adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position comprises the following steps:
and adding the adjusted similarity scores corresponding to the first type entity and the second type entity respectively to obtain the total score corresponding to the company position.
Optionally, the adjusting the similarity scores of the first type entity and the second type entity with the field information by using a preset anti-cheating adjustment algorithm according to the category information corresponding to the field information to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively, including:
if the category information corresponding to the field information is a working title, merging keywords repeatedly appearing in the working title, determining word number information corresponding to the merged working title, and adjusting the similarity score according to the word number information to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively;
if the category information corresponding to the field information is a work description, extracting keywords in the field information, and adjusting the similarity score according to the matching degree of the entity and the keywords to obtain adjusted similarity scores corresponding to the first category entity and the second category entity respectively;
And if the category information corresponding to the field information is wage information, determining an industry wage range in which the wage information is located, and adjusting the similarity score according to the industry wage range to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity.
Optionally, after the similarity scores corresponding to the first type of entity and the second type of entity are added to obtain a similarity total score corresponding to the company position, the method further includes:
calculating an industry hotness score, a company hotness score, a place hotness score and a position hotness score corresponding to the company position;
adding the industry hotness score, the company hotness score, the place hotness score and the job position hotness score to obtain a hotness total score corresponding to the company job position;
adding the heat total corresponding to the company position with the similarity total score to obtain an adjusted similarity total score corresponding to the company position;
displaying the company positions meeting the preset top ranking condition according to the similarity total score, including:
and sorting the company positions according to the adjusted similarity total score, and determining the company positions meeting the preset top ranking condition according to the sorting result to display.
Optionally, after the identifying the first type of entity and the second type of entity contained in the search statement, the method further comprises:
matching the first type entity and the second type entity with corresponding field information of each company respectively to obtain a matching result, and calculating a matching score of the search statement and each company according to the matching result;
and determining and displaying the company search results corresponding to the search sentences according to the matching scores.
Optionally, the matching the first type entity and the second type entity with the corresponding field information of each company respectively to obtain a matching result, and calculating the matching score of the search sentence and each company according to the matching result includes:
the first type entity and the second type entity are respectively and strictly matched with corresponding field information of a target company in each company;
if the first type entity and/or the second type entity are/is completely matched with the corresponding field information of the target company, calculating the matching score of the search statement and the target company according to the category information corresponding to the corresponding field information;
And if the first type entity and the second type entity are not completely matched with the corresponding field information of the target company, respectively carrying out fuzzy matching on the first type entity and the second type entity with the corresponding field information of the target company to obtain fuzzy matching scores of the first type entity and the second type entity with the corresponding field information, and determining the matching scores of the search statement and the target company according to the fuzzy matching scores.
Optionally, the calculating the matching score of the search sentence and the target company according to the category information corresponding to the corresponding field information includes:
if a plurality of field information is matched with the first type entity and/or the second type entity, respectively calculating strict matching scores of the plurality of field information and the first type entity and/or the second type entity by adopting a preset index weighting scoring algorithm according to category information respectively corresponding to the plurality of fields;
and adding the strict matching scores corresponding to the field information to obtain the matching scores of the search statement and the target company.
Optionally, the determining the matching score of the search sentence and the target company according to the fuzzy matching score includes:
And screening the maximum fuzzy matching score from the fuzzy matching scores, and determining the maximum fuzzy matching score as the matching score of the search sentence and the target company.
Optionally, determining and displaying the company search result corresponding to the search statement according to the matching score includes:
determining the maximum matching score in the matching scores of the search sentences and the companies, and judging whether the maximum matching score is a matching score obtained through strict matching according to the score magnitude corresponding to the maximum matching score;
if the maximum matching score is a matching score obtained through strict matching, judging whether target matching scores with the same score magnitude as the maximum matching score exist in the matching scores or not;
if the target matching score which has the same score magnitude as the maximum matching score does not exist, enhancing and displaying the company corresponding to the maximum matching score;
and if the target matching score with the same score magnitude as the maximum matching score exists, adding the companies respectively corresponding to the maximum matching score and the target matching score to a preset company list for display.
Optionally, after the determining whether the maximum matching score is a strict matching score according to the score magnitude corresponding to the maximum matching score, the method further includes:
if the maximum matching score is not the matching score obtained through strict matching, determining the number of companies with the matching score higher than a preset matching score;
if the number of the companies is equal to 1, enhancing and displaying the companies with the matching scores higher than the preset matching scores;
if the number of the companies is greater than 1, sequencing all the matching scores higher than the preset matching score from high to low, and judging whether the ratio between the matching score of the first ranking and the matching score of the second ranking is greater than or equal to the preset ratio according to the sequencing result;
if the matching score is larger than or equal to the preset ratio, enhancing and displaying the company corresponding to the first matching score;
if the matching score is smaller than the preset ratio, each company with the matching score higher than the preset matching score is added to a preset company list for display.
Optionally, displaying the company positions meeting the preset top ranking condition according to the similarity total score includes:
determining a target position provided by a company in the company search results;
Adding the similarity total score corresponding to the target position to obtain the added similarity total score corresponding to the target position;
and sorting all the positions according to the corresponding scores from high to low according to the added similarity total scores corresponding to the target positions and the similarity total scores corresponding to other positions, and determining the positions of the companies meeting the preset top ranking condition according to the sorting result to display.
According to a second aspect of the present invention, there is provided an information search apparatus comprising:
the acquisition unit is used for acquiring search sentences input by a user;
the identification unit is used for identifying the first type entity and the second type entity contained in the search statement;
the computing unit is used for acquiring a search list corresponding to the search statement, and respectively computing similarity scores of the first type entity and the second type entity and field information related to company positions in the search list;
the adding unit is used for adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position;
and the display unit is used for displaying the company positions meeting the preset ranking front condition according to the similarity total score.
Optionally, the apparatus further comprises: an adjusting unit is provided for adjusting the position of the adjusting unit,
the adjusting unit is configured to adjust similarity scores of the first type entity and the second type entity with the field information respectively by adopting a preset anti-cheating adjusting algorithm according to category information corresponding to the field information, so as to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively;
the adding unit is specifically configured to add the adjusted similarity scores corresponding to the first type entity and the second type entity respectively, so as to obtain a total score corresponding to the company position.
Optionally, the adjusting unit is specifically configured to, if the category information corresponding to the field information is a working title, combine keywords that repeatedly occur in the working title, determine word count information corresponding to the combined working title, and adjust the similarity score according to the word count information, so as to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively; if the category information corresponding to the field information is a work description, extracting keywords in the field information, and adjusting the similarity score according to the matching degree of the entity and the keywords to obtain adjusted similarity scores corresponding to the first category entity and the second category entity respectively; and if the category information corresponding to the field information is wage information, determining an industry wage range in which the wage information is located, and adjusting the similarity score according to the industry wage range to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity.
Optionally, the calculating unit is further configured to calculate an industry popularity score, a company popularity score, a place popularity score, and a job popularity score corresponding to the company job;
the adding unit is further configured to add the industry popularity score, the company popularity score, the place popularity score and the job popularity score to obtain a popularity total score corresponding to the company job;
the adding unit is further configured to add the heat total corresponding to the company position to the similarity total score to obtain an adjusted similarity total score corresponding to the company position;
the display unit is specifically configured to sort the company positions according to the adjusted similarity total score, and determine, according to the sorting result, the company positions that meet the preset top ranking condition for display.
Optionally, the computing unit is further configured to match the first type entity and the second type entity with corresponding field information of each company, obtain a matching result, and calculate a matching score of the search statement and each company according to the matching result;
and the display unit is also used for determining and displaying the company search results corresponding to the search sentences according to the matching scores.
Optionally, the computing unit includes: a strict matching module, a calculating module and a fuzzy matching module,
the strict matching module is used for strictly matching the first type entity and the second type entity with corresponding field information of target companies in the companies respectively;
the computing module is configured to compute a matching score of the search sentence and the target company according to category information corresponding to the corresponding field information if the first type entity and/or the second type entity are completely matched with the corresponding field information of the target company;
and the fuzzy matching module is used for respectively carrying out fuzzy matching on the first type entity and the second type entity and the corresponding field information of the target company if the first type entity and the second type entity are not completely matched with the corresponding field information of the target company, obtaining fuzzy matching scores of the first type entity and the second type entity and the corresponding field information, and determining the matching scores of the search statement and the target company according to the fuzzy matching scores.
Optionally, the computing module includes: a calculation sub-module and an addition sub-module,
The calculating submodule is used for respectively calculating the strict matching scores of the field information and the first-class entity and/or the second-class entity by adopting a preset index weighting scoring algorithm according to the class information respectively corresponding to the fields if the field information is matched with the first-class entity and/or the second-class entity;
and the adding sub-module is used for adding the strict matching scores corresponding to the field information to obtain the matching scores of the search statement and the target company.
Optionally, the fuzzy matching module is specifically configured to screen a maximum fuzzy matching score from the fuzzy matching scores, and determine the maximum fuzzy matching score as a matching score of the search sentence and the target company.
Optionally, the display unit includes: the judging module and the display module are used for judging whether the display module is in the display state,
the judging module is used for determining the maximum matching score in the matching scores of the search sentences and the companies, and judging whether the maximum matching score is a matching score obtained through strict matching according to the score magnitude corresponding to the maximum matching score;
the judging module is further configured to judge whether a target matching score with the same score magnitude as the maximum matching score exists in each matching score if the maximum matching score is a matching score obtained through strict matching;
The display module is used for carrying out enhanced display on the company corresponding to the maximum matching score if the target matching score which has the same score magnitude as the maximum matching score does not exist;
and the display module is further configured to, if there is a target matching score with the same score magnitude as the maximum matching score, add the companies corresponding to the maximum matching score and the target matching score to a preset company list for display.
Optionally, the display unit further includes: the determination module is configured to determine, based on the received data,
the determining module is used for determining the number of companies with the matching score higher than a preset matching score if the maximum matching score is not the matching score obtained through strict matching;
the display module is further used for carrying out enhanced display on the companies with the matching scores higher than the preset matching scores if the number of the companies is equal to 1;
the judging module is further configured to, if the number of companies is greater than 1, rank each matching score higher than a preset matching score from high to low, and judge whether a ratio between the matching score of the first rank and the matching score of the second rank is greater than or equal to a preset ratio according to a ranking result;
The display module is further used for carrying out enhanced display on the company corresponding to the first-ranked matching score if the first-ranked matching score is larger than or equal to a preset ratio;
and the display module is also used for adding each company with the matching score higher than the preset matching score to a preset company list for display if the matching score is smaller than the preset ratio.
Optionally, the display unit includes: a determining module, a adding and dividing module and a display module,
the determining module is used for determining a target position provided by a company in the company searching result;
the scoring module is used for scoring the similarity total score corresponding to the target position to obtain the scored similarity total score corresponding to the target position;
and the display module is used for sequencing the positions according to the corresponding scores from high to low according to the similarity total scores after the scoring corresponding to the target positions and the similarity total scores corresponding to other positions, and determining the positions of the companies meeting the preset top ranking condition according to the sequencing result to display.
According to a third aspect of the present invention, there is provided a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the above-described information search method.
According to a fourth aspect of the present invention, there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the above-mentioned information search method when executing the program.
Compared with the current mode of directly matching the search statement input by the user with the positions of different companies, the information searching method, device, storage medium and computer equipment provided by the invention can acquire the search statement input by the user; identifying a first type entity and a second type entity contained in the search statement; meanwhile, a search list corresponding to the search statement is obtained, and similarity scores of the first type entity and the second type entity and field information related to company positions in the search list are calculated respectively; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; finally, according to the similarity total score, the company positions meeting the preset ranking front condition are displayed, so that the similarity scores of the field information related to the first type entity and the second type entity and the company positions in the search sentences are calculated respectively through entity identification of the spoken or unstructured search sentences input by the user, the accurate matching of the search sentences input by the user and the different company positions can be realized, in addition, the different company positions are ordered through the similarity total score, the company positions meeting the preset ranking front condition are screened out and recommended to the user, and the company positions meeting the self requirements of the user can be recommended accurately for the user.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiments of the invention and together with the description serve to explain the invention and do not constitute a limitation on the invention. In the drawings:
fig. 1 shows a flowchart of an information searching method provided by an embodiment of the present invention;
FIG. 2 is a flowchart of another information searching method according to an embodiment of the present invention;
FIG. 3 illustrates an enhanced presentation schematic provided by an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an information searching apparatus according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of another information searching apparatus according to an embodiment of the present invention;
fig. 6 shows a schematic physical structure of a computer device according to an embodiment of the present invention.
Detailed Description
The invention will be described in detail hereinafter with reference to the drawings in conjunction with embodiments. It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other.
At present, the partial spoken or unstructured search sentences input by the user are not favorable for being matched with different company positions accurately, and the matching precision is low, so that the company positions meeting the requirements of the user cannot be provided for the user accurately.
In order to solve the above problem, an embodiment of the present invention provides an information searching method, as shown in fig. 1, including:
101. and acquiring a search statement input by a user.
The search statement input by the user is a partially spoken or unstructured search statement input by the user in the job-seeking website, for example, a manager who sleeps perfectly to be awake naturally. In order to overcome the defect that the prior art cannot directly match the partial spoken language or unstructured search statement input by the user with the company position accurately, the embodiment of the invention is mainly applicable to the search of the company position by carrying out entity identification on the search statement input by the user and calculating the similarity of the field information related to the company position and the entity contained in the search statement on the basis of the preliminary search result of the search engine, and can realize the accurate matching of the search statement input by the user with different company positions, thereby accurately recommending the company position meeting the self requirement for the user.
For the embodiment of the invention, when a user seeks a position, a corresponding search statement is input in a job-seeking website, and a search button is clicked to trigger a position search instruction to perform position search, a search engine provides a preliminary search result according to the search statement input by the user, an information search device can readjust and display the preliminary search result so as to improve search precision, and in the readjustment process, text error may exist in the search statement input by the user, for example, text error correction is performed on the search statement before entity identification is performed on the search statement, for example, "wake-up with nature" in the search statement is corrected to "sleep-wake-up with nature", and "in-line" in the search statement is corrected to "manager" so as to ensure that the entity contained in the search statement is accurately identified.
102. And identifying the first type entity and the second type entity contained in the search statement.
For the embodiment of the invention, after text correction is performed on the search statement input by a user, entity identification is performed on the search statement after the correction, particularly in the process of entity identification, word segmentation processing can be performed on the search statement by using a preset natural language model to obtain each word segment corresponding to the search statement, the preset natural language model can be a BERT natural language model, each word segment corresponding to the search statement is input into a preset entity identification model to perform entity identification, and the entity contained in the search statement is determined, wherein the preset entity identification model can be an LSTM network, the search statement at least contains one entity, particularly, each word segment corresponding to the search statement is input into the LSTM network, and then the entity category corresponding to each word segment can be determined according to the probability value, for example, the fact that the word segment corresponds to the entity corresponding to the natural language model is determined to be a sleep entity category corresponding to the natural language model, and the fact that the word segment corresponds to the corresponding to the entity category is not matched with the corresponding entity in the search statement can be determined by the corresponding to the position of the company, and the position is not accurately matched with the relevant information.
For the embodiment of the present invention, in order to ensure accuracy of entity recognition results, and to confirm real intention of a user, that is, whether to want to search for a position of a related company, before performing position search or position recommendation for the user, a company search may be performed according to an entity included in the search statement, and the company search results are displayed to the user, based on this, after the entity included in the search statement is identified, the method further includes: judging whether a target entity exists in the entities contained in the search statement; if the target entity exists, matching the field information corresponding to the target entity and different companies to obtain a matching result, and calculating the matching scores of the search statement and the different companies according to the matching result; and determining and displaying the company search results corresponding to the search sentences according to the matching scores. The search statement comprises at least one target entity in the entities, wherein the target entity can be a company entity and a product entity, and the field information corresponding to the company comprises a display name, an industry and commerce registration name, an alias, a product and the like of the company.
Specifically, in order to perform company searching according to entities in the search statement, the entities marked as companies in the search statement may be respectively matched with the display names, the business registration names and the alias field information of different companies, and simultaneously the entities marked as products in the search statement are matched with the product field information of different companies, so as to obtain matching results of the company entities and the product entities with related field information, and then according to the matching results, matching scores of the search statement and the different companies are calculated.
In a specific application scene, if a first type entity and a second type entity are simultaneously contained in a search sentence, namely, a company entity and a product entity are simultaneously contained in the search sentence, respectively matching the first type entity and the second type entity with corresponding field information of each company to obtain a matching result, and calculating a matching score of the search sentence and each company according to the matching result; and determining and displaying the company search results corresponding to the search sentences according to the matching scores.
For the process of calculating the matching score, the matching the target entity with the field information corresponding to different companies to obtain a matching result, and calculating the matching score of the search statement and the different companies according to the matching result, including: carrying out strict matching on the target entity and field information corresponding to a target company in each company; if the target entity is completely matched with the field information of the target company, calculating the matching score of the search statement and the target company according to the category information corresponding to the field information; if the field information of the target entity and the field information of the target company are not completely matched, fuzzy matching is carried out on the field information of the target entity and the field information of the target company, fuzzy matching scores of the target entity and the field information are obtained, and the matching scores of the search statement and the target company are determined according to the fuzzy matching scores.
Further, the calculating the matching score of the search sentence and the target company according to the category information corresponding to the field information includes: if a plurality of field information is matched with the target entity, respectively calculating strict matching scores of the plurality of field information and the target entity by adopting a preset index weighting scoring algorithm according to category information respectively corresponding to the plurality of fields; and adding the field information and the strict matching scores of the target entity to obtain the matching scores of the search statement and the target company. Further, the fuzzy matching of the target entity and the field information of the target company to obtain a fuzzy matching score of the target entity and the field information, and determining the matching score of the search sentence and the target company according to the fuzzy matching score, includes: if the target company corresponds to a plurality of field information, performing fuzzy matching on the target entity and the plurality of field information respectively, and determining fuzzy matching scores of the target entity and the plurality of field information respectively; and screening out the maximum fuzzy matching score from the fuzzy matching scores, and determining the maximum fuzzy matching score as the matching score of the search statement and the target company. The target company is any company recorded in a preset company record table, and field information of different companies is recorded in the preset company record table.
In a specific application scene, if a first type entity and a second type entity are contained in a search sentence at the same time, namely, a company entity and a product entity are contained in the search sentence at the same time, the first type entity and the second type entity are respectively strictly matched with corresponding field information of a target company in each company; if the first type entity and/or the second type entity are/is completely matched with the corresponding field information of the target company, calculating the matching score of the search statement and the target company according to the category information corresponding to the corresponding field information; and if the first type entity and the second type entity are not completely matched with the corresponding field information of the target company, respectively carrying out fuzzy matching on the first type entity and the second type entity with the corresponding field information of the target company to obtain fuzzy matching scores of the first type entity and the second type entity with the corresponding field information, and determining the matching scores of the search statement and the target company according to the fuzzy matching scores.
Further, the calculating the matching score of the search sentence and the target company according to the category information corresponding to the corresponding field information includes: if a plurality of field information is matched with the first type entity and/or the second type entity, respectively calculating strict matching scores of the plurality of field information and the first type entity and/or the second type entity by adopting a preset index weighting scoring algorithm according to category information respectively corresponding to the plurality of fields; and adding the strict matching scores corresponding to the field information to obtain the matching scores of the search statement and the target company. Further, the determining the matching score of the search sentence and the target company according to the fuzzy matching score includes: and screening the maximum fuzzy matching score from the fuzzy matching scores, and determining the maximum fuzzy matching score as the matching score of the search sentence and the target company.
Specifically, the company entity in the search statement is strictly matched with the display name, the business registration name and the alias field information of the target company respectively, and the product entity in the search statement is strictly matched with the product field information of the target company at the same time, if the target entity in the search statement is strictly matched with the field of the target companyIf the information is completely consistent, the two are completely matched, corresponding score levels are determined according to the category information corresponding to the matched field information, and the matching scores of the target entity and the field information are determined according to the score levels, for example, if the company entity in the search statement is completely matched with the display name of the target company, the matching score of the company entity and the display name of the target company is determined to be 10 6 Dividing; if the company entity in the search statement is completely matched with the business registration name of the target company, determining the matching score of the company entity and the business registration name of the target company to be 10 5 Dividing; if the company entity and the alias of the target company in the search statement are completely matched, determining that the matching score of the company entity and the alias of the target company is 10 4 Dividing; if the product entity in the search statement is completely matched with the product of the target company, determining that the matching score of the product entity and the product of the target company is 10 3 Dividing into two parts. Further, if in the course of the strict match, if the plurality of field information is completely matched, the plurality of field information is added to the matching score of the target entity to obtain a matching score for searching the search sentence and the target company, and immediately above example, if the company entity is completely matched with the display name and the business registration name of the target company and the product entity is completely matched with the product of the target company, the matching score of the search sentence and the target company is determined to be 10 6 +10 5 +10 3 According to the matching scores of different quantity levels, the matching score between the target entity and the field information can be determined by selecting corresponding score levels according to the types of the completely matched field information by adopting a preset index weighting scoring algorithm, and further, according to the matching scores of different quantity levels, the completely matched field information of the target entity can be directly obtained, for example, if the matching score between the search statement and the target company is more than 10 6 And dividing the description search statement into a complete match with the display name of the target company.
Further, if the company entity and the product entity in the search statement are not completely matched with the display name, the business registration name, the alias and the product field information of the target company, the company entity and the product entity are respectively and fuzzy matched with the display name, the business registration name, the alias and the product field information of the target company, the fuzzy matching score of the field information of the target entity and the target company is determined by adopting a score level lower than that of strict matching, so that whether the entity in the search statement is fuzzy matched with the field information of the target company or strictly matched can be determined according to the score level corresponding to the matching score, for example, the fuzzy matching score of the display name, the business registration name and the alias of the company entity and the target company is 40, 60 and 20 respectively, the fuzzy matching score of the product entity and the product of the target company is 30, then the highest score of 60 is selected as the matching score of the search statement and the target company, if the highest score is lower than a preset score, for example, the probability of matching the search statement and the target company is extremely small can be directly discarded.
It should be noted that, because the fuzzy matching mode and the strict matching adopt different score magnitudes to determine the matching scores between the field information corresponding to the target entity and the target company, the matching score difference calculated by the two modes is increased, so that orthogonality can be kept, no mutual influence can be generated, and further, the fuzzy matching and the strict matching can be performed simultaneously in the searching process of the first company, and the matching scores calculated by the two modes are added to finally obtain the matching scores of the search statement and the target company.
Further, after calculating the matching scores of the search sentence and different companies through fuzzy matching and strict matching, screening out the companies meeting the self requirements of the user for display, as an optional embodiment, determining and displaying the company search results corresponding to the search sentence according to the matching scores, including: determining the maximum matching score in the matching scores of the search sentences and the companies, and judging whether the maximum matching score is a matching score obtained through strict matching according to the score magnitude corresponding to the maximum matching score; if the maximum matching score is a matching score obtained through strict matching, judging whether target matching scores with the same score magnitude as the maximum matching score exist in the matching scores or not; if the target matching score which has the same score magnitude as the maximum matching score does not exist, enhancing and displaying the company corresponding to the maximum matching score; and if the target matching score with the same score magnitude as the maximum matching score exists, adding the companies respectively corresponding to the maximum matching score and the target matching score to a preset company list for display.
For example, a maximum match score of 10 is determined 5 +10 4 Score corresponding to a score of magnitude of 10 5 The score can be determined to be the score obtained by strict matching according to the score magnitude corresponding to the maximum matching score, and further, whether the score magnitude is 10 is still present in the score of the search statement corresponding to different companies or not is judged 5 If the target matching score does not exist, the company is the only company which meets the requirement of the user, and the matching degree with the search statement input by the user is high, so that the company corresponding to the maximum matching score is enhanced and displayed, namely the information of the company is displayed in a single display board above a work list, as shown in fig. 3; if the target matching score with the same score magnitude as the maximum matching score also exists in each matching score, the method indicates that a plurality of companies meet the requirements of the user and have high matching degree with search sentences input by the user, so that the companies corresponding to the maximum matching score and the target matching score are added to a preset company list together and are displayed to the user in a sorting mode according to the score.
Further, after the determining whether the maximum matching score is a strict matching score according to the score magnitude corresponding to the maximum matching score, the method further includes: if the maximum matching score is not the matching score obtained through strict matching, determining the number of companies with the matching score higher than a preset matching score; if the number of the companies is equal to 1, enhancing and displaying the companies with the matching scores higher than the preset matching scores; if the number of the companies is greater than 1, sequencing all the matching scores higher than the preset matching score from high to low, and judging whether the ratio between the matching score of the first ranking and the matching score of the second ranking is greater than or equal to the preset ratio according to the sequencing result; if the matching score is larger than or equal to the preset ratio, enhancing and displaying the company corresponding to the first matching score; if the matching score is smaller than the preset ratio, each company with the matching score higher than the preset matching score is added to a preset company list for display. The preset matching score and the preset ratio can be set according to service requirements, for example, the matching score is set to be 10 minutes, and the preset ratio is set to be 1.5.
For example, the maximum match score is determined to be 70 points, which corresponds to a score level that is less than the minimum score level of 10 for a strict match 3 The maximum matching score is a matching score obtained through fuzzy matching, further, the number of companies with the matching score higher than a preset matching score is determined, if the number of the companies with the matching score higher than the preset matching score is 1, only one company is indicated to possibly meet the requirement of a user, and the company is enhanced and displayed; if the number of the companies with the matching scores higher than the preset matching scores is greater than 1, the situation that a plurality of companies possibly meet the requirements of users is indicated, further, the companies with the matching scores higher than the preset matching scores are ranked according to the scores, the ratio between the matching scores corresponding to the first ranked company and the matching scores corresponding to the second ranked company is calculated, if the ratio is greater than or equal to the preset ratio, the situation that the first ranked company, namely the company with the highest matching score, meets the requirements of the users more than other companies is indicated, and therefore enhancement display is carried out; if the ratio is smaller than the preset ratio, the matching degree of the companies with the matching scores higher than the preset matching score and the search statement input by the user is not very different, and the requirements of the user can be met, so that the companies with the matching scores higher than the preset matching score are added to a preset company list and are displayed to the user after being ranked according to the matching score, on the basis, the matching scores of all the companies in the company list obtained through fuzzy matching can be adjusted according to the number of the company positions and the company scale, all the companies in the company list can be ranked again based on the adjusted matching scores, and the ranking results are displayed to the user, for example, the public The larger the size of the company or the larger the number of the provided positions, the larger the matching score is adjusted upwards, so that the company which is most beneficial to the user is preferentially displayed to the user in the company search result, and the probability of successful job hunting of the user is improved.
By carrying out strict matching and fuzzy matching on the entities contained in the search statement and field information related to different companies, the corresponding companies can be screened and displayed to the user, so that the user can confirm whether to search the positions of related companies in a preset company list, namely, the real intention of the user can be confirmed, and meanwhile, whether to accurately identify the entities of the search statement or not, for example, whether to identify the company entities in the search statement as other entities or not can be determined according to the search result of the company.
103. And obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list.
The method comprises the steps that a search list is a preliminary search result obtained by a search engine according to a search sentence of a user, the search list comprises different company positions and field information corresponding to the different company positions, the field information of the company positions comprises field information such as titles, details, company names, addresses, skills, welfare, industries and products, and the like; if the position which the user wants to search is not limited to the position provided by the relevant company in the company search result, searching all the company positions recorded in the search list and feeding back the search result to the user, wherein the field information corresponding to different company positions is recorded in the search list.
In a specific application scenario, if the search statement includes both the first type entity and the second type entity, similarity scores of field information corresponding to different company positions in the search list between the first type entity and the second type entity are calculated respectively, so that similarity total scores corresponding to different company positions are calculated according to the similarity scores corresponding to the first type entity and the second type entity respectively.
In addition, if the positions which the user wants to search are not limited to the positions provided by the relevant companies in the company search results, after calculating the similarity scores of the field information of the entities and the positions of different companies, the positions provided by the relevant companies in the company search results can be subjected to the adding processing based on the similarity scores, and the positions provided by the relevant companies in the company search results and the positions provided by the relevant companies in the user search sentences are subjected to the adding processing because the matching degree of the relevant companies in the company search results and the entities in the user search sentences is higher, so that the positions provided by the relevant companies can be preferentially displayed to the user, and the search requirements of the user can be met more easily.
104. And adding the similarity scores corresponding to the first type of entity and the second type of entity respectively to obtain a similarity total score corresponding to the company position.
For the embodiment of the invention, if a plurality of entities are contained in a search sentence, the similarity scores of the related field information of the plurality of entities and the company positions are added to obtain the similarity total score of the company positions in the search sentence and the search list input by a user, and in a specific application scene, if the first type entity and the second type entity are simultaneously contained in the search sentence, the similarity scores of the related field information of the first type entity and the second type entity and the company positions are respectively added to obtain the similarity total score of the company positions in the search sentence and the search list input by the user, so that the company positions in the search list are sequenced and displayed according to the similarity total score.
105. And displaying the company positions meeting the preset top ranking conditions according to the similarity total score.
For the embodiment of the invention, after calculating the similarity scores of the field information related to the entity and the company positions, sorting the company positions from high to low according to the similarity scores, screening the company positions meeting the preset top ranking condition according to the sorting result, and displaying the company positions to the user.
Compared with the current mode of directly matching the search statement input by the user with the positions of different companies, the information search method provided by the embodiment of the invention can acquire the search statement input by the user; identifying a first type entity and a second type entity contained in the search statement; meanwhile, a search list corresponding to the search statement is obtained, and similarity scores of the first type entity and the second type entity and field information related to company positions in the search list are calculated respectively; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; finally, according to the similarity total score, the company positions meeting the preset ranking front condition are displayed, so that the similarity scores of the field information related to the first type entity and the second type entity and the company positions in the search sentences are calculated respectively through entity identification of the spoken or unstructured search sentences input by the user, the accurate matching of the search sentences input by the user and the different company positions can be realized, in addition, the different company positions are ordered through the similarity total score, the company positions meeting the preset ranking front condition are screened out and recommended to the user, and the company positions meeting the self requirements of the user can be recommended accurately for the user.
Further, in order to better illustrate the searching process of the company position, as a refinement and extension of the foregoing embodiment, an embodiment of the present invention provides another information searching method, as shown in fig. 2, where the method includes:
201. and acquiring a search statement input by a user.
For the embodiment of the invention, when a user searches for positions on a job-seeking website, search sentences input by the user are obtained, most of the search sentences are spoken in a biased way or are unstructured, entity identification is needed for the search sentences so as to be capable of accurately matching the search sentences with positions of different companies, and the similarity of field information related to the positions of the companies and entities contained in the search sentences is calculated, so that the accurate matching of the search sentences input by the user and the positions of the different companies is realized.
202. And identifying the first type entity and the second type entity contained in the search statement.
For the embodiment of the present invention, in order to perform entity identification on a search statement input by a user, step 202 specifically includes: performing word segmentation processing on the search sentence to obtain each word segmentation corresponding to the search sentence; and inputting each word segment into a preset entity recognition model to perform entity recognition, and determining the entity contained in the search statement. Further, the preset entity recognition model includes a first recurrent neural network and a second recurrent neural network, the inputting the individual segmentation words into the preset entity recognition model performs entity recognition, and determining the entity included in the search statement includes: inputting each word segment into a first recurrent neural network according to the sequence of the word segment in a search sentence to perform feature extraction, and obtaining a first feature vector corresponding to each word segment; inputting the segmented words into a second recurrent neural network in reverse order according to the sequence of the segmented words in the search statement to perform feature extraction, and obtaining second feature vectors corresponding to the segmented words; and combining the first feature vector and the second feature vector to obtain a combined feature vector, and determining the entity category corresponding to each word according to the combined feature vector. Further, the preset entity recognition model further includes a conditional random field network, and the determining, according to the combined feature vector, the entity category corresponding to each word segment includes: determining probability values of the different entity categories of the segmented words according to the combined feature vectors; correcting the probability values of the different entity categories of the segmented words by using the conditional random network to obtain corrected probability values of the different entity categories of the segmented words; and determining the entity category corresponding to each word segment based on the probability value that each word segment after correction belongs to different entity categories.
Specifically, firstly, word segmentation processing is performed on a search sentence input by a user by using a preset natural language model, each word segment corresponding to the search sentence is obtained, each word segment is converted into a corresponding word vector, wherein the preset natural language model can be but is not limited to a BERT natural language model, then each word segment corresponding to the search sentence is input into a preset entity recognition model for entity recognition, an entity contained in the search sentence is determined, the entity recognition model comprises a first recurrent neural network and a second recurrent neural network, a bidirectional LSTM network is formed by the first recurrent neural network and the second recurrent neural network, and when the entity recognition is performed, the word vectors corresponding to the segmented words are input into a first recurrent neural network according to the sequence of the segmented words in a search sentence to perform feature extraction, a first feature vector corresponding to the segmented words is obtained, the word vectors corresponding to the segmented words are input into a second recurrent neural network according to the sequence of the segmented words in the search sentence to perform feature extraction, a second feature vector corresponding to the segmented words is obtained, the first feature vector and the second feature vector are combined to obtain combined feature vectors, for example, the two 32-dimensional feature vectors are spliced into 64-dimensional feature vectors, and the combined feature vectors are input into a softmax network to obtain probability values of the segmented words belonging to different entity categories.
For the embodiment of the invention, the entities can be further classified according to the location information of the entities, such as B-Organization, I-Organization, wherein 'B-' represents the initial location of the entity, 'I-' represents the middle location of the entity, and 'Organization' represents the category of the entity, if 8 types of entities are involved, including company, industry, position, skill, welfare, wage, product and place, and other categories are added, each word corresponds to a vector with the length of 17, all values of the vector are added to be equal to 1, then the maximum value in the probability value is selected, and the category of the entity corresponding to the maximum value is determined according to the category of the entity corresponding to the word. Further, in order to improve accuracy of entity identification, constraint is performed on an output result of the softmax network by using a conditional random network CRF, for example, a beginning of each entity must be "B-", "B-ORG" cannot be followed by "I-LOC", so that probability values of each word belonging to different entity categories are corrected by using the CRF network, and entity categories corresponding to each word are determined based on the corrected probability values of each word belonging to different entity categories.
203. And obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list.
For the embodiment of the invention, a preset text similarity algorithm is utilized to calculate similarity scores of the field information related to the company position, as shown in table 1, similarity scores of the field information related to the company position, such as a working entity and the title, the details, the company name, the address, the skill, the welfare, the industry, the product and the like are calculated respectively, if other entities, such as the company, the industry, the place and the like still exist in a search sentence, the similarity scores of the other entities and the field information, such as the title, the details, the company name, the address, the skill, the welfare, the industry, the product and the like are calculated continuously, and further, if a plurality of similarity scores exist between the entities and different field information, the maximum similarity score is selected as the similarity score corresponding to the company position entity, such as the similarity score corresponding to the company entity with the title, the details, the company name is 5 score, the company name is 3 score and the company name is 10 score, the similarity score corresponding to the company position entity is determined to be 10 score, the fact that the corresponding to the company position entity is prevented from repeatedly appearing in the field information, if the second category is not repeated, the first category is displayed, and the first category score is determined according to the total score is obtained, and the total score is ranked according to the accuracy, and the first category score is obtained.
Further, for a specific calculation process of similarity scores of field information related to the company position by the entity, a preset text similarity algorithm is adopted to calculate word frequency of the entity in the field information of the company position and weight corresponding to the entity, and the similarity scores of title field information of the working entity and the company position are calculated as examples, and the specific formula is as follows:
wherein, different company positions correspond to different documents, TF represents word frequency, namely frequency of occurrence of working entities in header field information, IDF represents weight corresponding to the working entities, N represents number of documents containing the working entities in the header, N represents total number of documents of the company positions, dl represents length of the header field information, dl avg Representing the average length, k, of header field information 1 And b represents an optimization parameter, typically empirically set, k 1 2, b is 0.75, and then multiplying the calculated word frequency TF by the weight IDF corresponding to the working entity to obtain the similarity evaluation of the title field information of the working entity and the position of a certain companyThe similarity scores of the entities contained in the search statement and the relevant field information of different company positions can be calculated according to the method.
204. And respectively adjusting the similarity scores of the first type entity and the second type entity and the field information by adopting a preset anti-cheating adjustment algorithm according to the category information corresponding to the field information, so as to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity.
For the embodiment of the invention, the position provider can perform certain cheating on the position title or descriptive content to improve the correlation degree between the position provider and the search statement, so the embodiment of the invention adopts an anti-cheating adjustment algorithm to perform downward adjustment on the similarity scores of certain entities and field information, thereby leading to relatively back similarity ranking of the position provider with suspected cheating, and further effectively preventing the position provider from cheating.
In a specific application scenario, according to category information corresponding to the field information, a preset anti-cheating adjustment algorithm is adopted to adjust similarity scores of the entity and the field information, and the obtained similarity scores after adjustment of the entity and the field information comprise: if the category information corresponding to the field information is a working title, merging keywords repeatedly appearing in the working title, determining word number information corresponding to the merged working title, and adjusting the similarity score according to the word number information to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively; if the category information corresponding to the field information is a work description, extracting keywords in the field information, and adjusting the similarity score according to the matching degree of the entity and the keywords to obtain adjusted similarity scores corresponding to the first category entity and the second category entity respectively; and if the category information corresponding to the field information is wage information, determining an industry wage range in which the wage information is located, and adjusting the similarity score according to the industry wage range to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity.
Specifically, after calculating the similarity score of the related field information of the entity and different company positions, if the field information is a working title, because most of the entity contained in the search statement appears in the working title, if the job provider adds a large number of keywords in the working title, a longer title is created, so that the title is easier to match with the entity, the word frequency of the occurrence of the keywords in the working title needs to be considered, after the repeated keywords are combined, the word number corresponding to the working title is counted again, thereby effectively preventing the job provider from cheating, in addition, because the field length item in the text similarity algorithm per se is attenuated, the short field can be subjected to corresponding score addition once being matched with the entity, but for example, "sales" is not more matched with "sales" than "when the field information is already reduced to a certain extent, the additional score addition is not given to the field length any more, in order to ensure that the working requirement is correctly expressed, on the contrary, the score of the excessively short field is attenuated, based on the fact, in the embodiment of the invention, the first title is set to be equal to the preset length of the first title or the linear score of the first title is set to the first title, if the length of the preset length of the first title is equal to the linear score of the first title is set to the corresponding length of the first title or the corresponding length of the job title is set to the first title; if the title length of the position of a certain company is larger than the second preset title length, after calculating the similarity score of the entity and the title field information, carrying out corresponding logarithmic function attenuation on the similarity score, thereby effectively preventing the position provider from obtaining higher correlation by adding keywords in the title.
Further, if the field information is a job description, since the job description is usually a long text, a company job provider may repeatedly place some important keywords in the job description to obtain a higher degree of correlation, and for this case, a summary extraction technology is used to extract a preset number of keywords in the job description, and in the process of matching an entity with the description field information, if the entity matches other words besides the keywords in the job description, the calculated similarity score is subjected to score attenuation, and in the process of matching the entity with the description field information, if the entity matches the keywords in the job description field information repeatedly, the calculated similarity score is also subjected to score attenuation, so that the job provider is effectively prevented from cheating by adding the keywords in the job description, and a higher degree of correlation is obtained.
Further, if the field information is payroll, since the company position provider can easily obtain higher correlation by providing too high payroll, in order to effectively prevent the cheating situation, a corresponding industry payroll range can be set, whether the payroll is within the industry payroll range of the company position is judged based on the industry to which the company position belongs, if not, the payroll of the position provider can be considered to be abnormal, similarity scores may need to be attenuated, specifically, the industry in which the company position is located is first selected, the payroll of all the positions of the industry is obtained, and ranked according to the height of the payroll, then the corresponding percentage of the ranked payroll is obtained, the industry payroll range is determined, for example, 100 payroll ranges are respectively ranked by 1%,5%,95%,99%, the industry payroll range is determined, namely, if the payroll of the company position is within the range of 5% to 95%, the normal payroll of the company position is determined, and similarity scores are calculated according to the similarity scores; if the wages of the certain company position are in the range of 1% to 5% and 95% to 99%, determining that the wages of the certain company position are in a flat area, not adjusting the calculated similarity score, and if the wages of the certain company position are less than 1% or more than 99%, determining that the wages of the certain company position are abnormal wages, and carrying out attenuation treatment on the calculated similarity score.
205. And adding the adjusted similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position.
For the embodiment of the invention, after calculating the similarity scores of the field information related to the company position of the entity, if the search statement contains a plurality of entities, the similarity scores of the field information related to the company position of the plurality of entities are added to obtain the total score corresponding to the company position, as shown in table 1, the similarity score corresponding to the working entity is 6 points, the similarity score corresponding to the company entity is 10 points, the similarity score corresponding to the industry entity is 9 points, the similarity score corresponding to the place entity is 9 points, the similarity score corresponding to the skill entity is 16 points, the similarity score corresponding to the product entity is 10 points, the similarity score corresponding to the welfare entity is 6 points, and the similarity scores corresponding to the entities are added to 6+10+9+9+16+10+6=66 points. Further, if the similarity scores of the entities and the field information are adjusted by adopting an anti-cheating adjustment algorithm, the similarity scores of the entities and the field information after adjustment are added to obtain total scores corresponding to the company positions, so that the total scores corresponding to different company positions can be calculated according to the mode.
Further, in the actual searching experience, the user often wants to see the work of a large company and the hotter industry, if a large number of users can be obtained, the heat distribution of the work can be determined based on the interactive behavior of the users, so as to provide the users with searching sequences more conforming to market trend, however, if in a cold start scene, because of lack of user data, the user needs to start from information of different company positions, possible heat information is mined, and in order to mine the heat information, four heat indexes are defined in the embodiment of the invention, including: the method comprises the steps of determining total heat scores corresponding to different company positions based on the four indexes, and adjusting the total similarity scores corresponding to different positions based on the total heat scores so as to be capable of preferentially displaying the company positions meeting market demands or having higher heat to a user after sorting according to scores, and adding the similarity scores of field information related to the company positions of the entities to obtain the total scores corresponding to the company positions, wherein the method further comprises the steps of: calculating an industry hotness score, a company hotness score, a place hotness score and a position hotness score corresponding to the company position; adding the industry hotness score, the company hotness score, the place hotness score and the job position hotness score to obtain a hotness total score corresponding to the company job position; and adding the heat total corresponding to the company position with the similarity total score to obtain the adjusted similarity total score corresponding to the company position.
Specifically, the calculation formulas of the industry popularity score, the company popularity score, the place popularity score and the job popularity score are as follows:
wherein HEAT ind Representative industry hotness score, N max Represents the maximum value of the number of industry companies, SAL represents industry wage mean, SAL max Representing the maximum value of industry payroll means.
Wherein HEAT com Represents the corporate hotness score, N max Representing the maximum of the number of company release positions, SIZE representing the company SIZE max Representing the maximum of the company scale.
Wherein HEAT job Representing job hotness score, day representing days since job release to date, SAL representing payroll of company job, SAL max Representing the maximum value of payroll in the industry corresponding to the company position.
Wherein HEAT city Representative site hotness score, N max Representing the maximum of the number of jobs contained in all cities.
According to the formula, the industry popularity score, the company popularity score, the place popularity score and the position popularity score corresponding to different company positions can be calculated, the industry popularity score, the company popularity score, the place popularity score and the position popularity score are added to obtain the popularity total score corresponding to the different company positions, and then the popularity total score and the similarity total score corresponding to the different company positions are added to obtain the adjusted total score corresponding to the different company positions, so that the different company positions are ordered according to the adjusted total score, and the company positions which meet the own demands of users and have market popularity are displayed to the users in priority.
206. And displaying the company positions meeting the preset top ranking conditions according to the similarity total score.
For the embodiment of the present invention, in order to provide a search ranking more in line with market trend for a user, the ranking the company positions according to the total score, and determining, according to a ranking result, that the company positions in line with a preset top ranking condition are displayed, includes: and sorting the company positions according to the adjusted similarity total score, and determining the company positions meeting the preset top ranking condition according to the sorting result to display.
Further, if the company search is performed before the work search, the positions provided by the related companies in the company search results may be subjected to a scoring process based on the similarity total score, so that the positions provided by the companies with higher relevance in the company search results are preferentially displayed to the user, so as to meet the search requirement of the user, based on this, the company positions are ranked according to the total score, and the company positions meeting the preset pre-ranking condition are determined according to the ranking result to be displayed, including: determining a target position provided by a company in the company search results; adding the similarity total score corresponding to the target position to obtain the added similarity total score corresponding to the target position; and sorting the positions according to the corresponding scores from high to low according to the similarity total scores of the target positions after the scoring and the similarity total scores of other positions, and determining the positions of the companies meeting the preset top ranking condition according to the sorting result to display. The target position is a company position provided by a relevant company in the company search result.
For example, the company search results include company a and company B, after the similarity total scores corresponding to different company positions are calculated, the target positions provided by company a and company B are determined, and the similarity total scores corresponding to the target positions are subjected to adding processing.
Compared with the current mode of directly matching the search statement input by the user with the positions of different companies, the information searching method provided by the embodiment of the invention can acquire the search statement input by the user; identifying a first type entity and a second type entity contained in the search statement; meanwhile, a search list corresponding to the search statement is obtained, and similarity scores of the first type entity and the second type entity and field information related to company positions in the search list are calculated respectively; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; finally, according to the similarity total score, the company positions meeting the preset ranking front condition are displayed, so that the similarity scores of the field information related to the first type entity and the second type entity and the company positions in the search sentences are calculated respectively through entity identification of the spoken or unstructured search sentences input by the user, the accurate matching of the search sentences input by the user and the different company positions can be realized, in addition, the different company positions are ordered through the similarity total score, the company positions meeting the preset ranking front condition are screened out and recommended to the user, and the company positions meeting the self requirements of the user can be recommended accurately for the user.
Further, as a specific implementation of fig. 1, an embodiment of the present invention provides an information searching apparatus, as shown in fig. 4, where the apparatus includes: an acquisition unit 31, an identification unit 32, a calculation unit 33, an addition unit 34, and a presentation unit 35.
The obtaining unit 31 may be configured to obtain a search term input by a user. The obtaining unit 31 is a main functional module for obtaining a search term input by a user in the present apparatus.
The identifying unit 32 may be configured to identify and identify a first type of entity and a second type of entity included in the search term. The identifying unit 32 is a main functional module for identifying the first type of entity and the second type of entity contained in the search statement in the present device, and is also a core module.
The calculating unit 33 may be configured to obtain a search list corresponding to the search sentence, and calculate similarity scores of the first type entity and the second type entity and field information related to company positions in the search list respectively. The calculating unit 33 is a main functional module, and is also a core module, for obtaining a search list corresponding to the search sentence in the device, and calculating similarity scores of the first type of entity and the second type of entity and field information related to company positions in the search list respectively.
The adding unit 34 may be configured to add the similarity scores corresponding to the first type entity and the second type entity respectively, to obtain a similarity total score corresponding to the company position. The adding unit 34 is a main functional module in the present apparatus, and adds the similarity scores corresponding to the first type of entity and the second type of entity respectively, so as to obtain a similarity total score corresponding to the company position.
The display unit 35 may be configured to display the company positions meeting the preset top ranking condition according to the similarity total score. The display unit 35 is a main functional module for displaying the company positions meeting the preset top ranking condition according to the similarity total score in the device, and is also a core module.
For the embodiment of the present invention, as shown in fig. 5, in order to prevent cheating of the job provider, the calculated similarity score needs to be adjusted, and based on this, the apparatus further includes: an adjustment unit 36.
The adjusting unit 36 may be configured to adjust similarity scores of the first type entity and the second type entity with the field information by using a preset anti-cheating adjustment algorithm according to the category information corresponding to the field information, so as to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively.
The adding unit 34 may be specifically configured to add the adjusted similarity scores corresponding to the first type of entity and the second type of entity respectively, so as to obtain a similarity total score corresponding to the company position.
In a specific application scenario, the adjusting unit 36 may be specifically configured to, if the category information corresponding to the field information is a working title, combine keywords that repeatedly occur in the working title, determine word count information corresponding to the combined working title, and adjust the similarity score according to the word count information, so as to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively; if the category information corresponding to the field information is a work description, extracting keywords in the field information, and adjusting the similarity score according to the matching degree of the entity and the keywords to obtain adjusted similarity scores corresponding to the first category entity and the second category entity respectively; and if the category information corresponding to the field information is wage information, determining an industry wage range in which the wage information is located, and adjusting the similarity score according to the industry wage range to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity.
Further, the calculating unit 33 may be further configured to calculate an industry popularity score, a company popularity score, a place popularity score, and a job popularity score corresponding to the company job.
The adding unit 34 may be further configured to add the industry popularity score, the company popularity score, the place popularity score, and the job popularity score to obtain a popularity total score corresponding to the company job.
The adding unit 34 may be further configured to add the total heat corresponding to the company position to the total similarity score to obtain an adjusted total similarity score corresponding to the company position.
The display unit 35 may be specifically configured to sort the company positions according to the adjusted similarity total score, and determine, according to the sorting result, the company positions that meet the preset top ranking condition for display.
In a specific application scenario, in order to calculate a matching score of the search term and the target company, the calculating unit 33 includes: a strict match module 331, a calculation module 332, and a fuzzy match module 333.
The strict matching module 331 may be configured to strictly match the first type of entity and the second type of entity with corresponding field information of a target company in the respective companies, respectively.
The calculating module 332 may be configured to calculate, if the first type entity and/or the second type entity are completely matched with the corresponding field information of the target company, a matching score of the search sentence and the target company according to the category information corresponding to the corresponding field information.
The fuzzy matching module 333 may be configured to, if the first type entity and the second type entity are not completely matched with the corresponding field information of the target company, perform fuzzy matching on the first type entity and the second type entity with the corresponding field information of the target company, obtain fuzzy matching scores of the first type entity and the second type entity with the corresponding field information, and determine a matching score of the search sentence and the target company according to the fuzzy matching scores.
Further, to determine the matching score of the search term to the target company, the calculation module 332 includes a calculation sub-module and an addition sub-module.
The calculating submodule can be used for respectively calculating the strict matching scores of the field information and the first-class entity and/or the second-class entity by adopting a preset exponential weighting scoring algorithm according to the class information respectively corresponding to the field information if the field information is matched with the first-class entity and/or the second-class entity.
And the addition sub-module can be used for adding the strict matching scores corresponding to the field information to obtain the matching scores of the search statement and the target company.
In a specific application scenario, the fuzzy matching module 333 may be specifically configured to screen out a maximum fuzzy matching score from the fuzzy matching scores, and determine the maximum fuzzy matching score as a matching score of the search sentence and the target company.
Further, in order to display the company search results, the display unit 35 includes a judgment module 351 and a display module 352.
The determining module 351 may be configured to determine a maximum matching score of the matching scores of the search sentence and the companies, and determine, according to a score level corresponding to the maximum matching score, whether the maximum matching score is a matching score obtained by strict matching.
The determining module 351 may be further configured to determine whether a target matching score having the same score magnitude as the maximum matching score exists in each matching score if the maximum matching score is a matching score obtained by strict matching.
The display module 352 may be configured to perform enhanced display on a company corresponding to the maximum matching score if there is no target matching score having the same score magnitude as the maximum matching score.
The display module 352 may be further configured to, if there is a target matching score having the same score magnitude as the maximum matching score, add the company corresponding to the maximum matching score and the target matching score to a preset company list for display.
Further, the display unit 35 performs enhanced display on the company search results, and further includes: a determination module 353.
The determining module 353 may be configured to determine the number of companies with the matching score higher than the preset matching score if the maximum matching score is not the matching score obtained by strict matching.
The display module 352 may be further configured to enhance and display a company with a matching score higher than a preset matching score if the number of companies is equal to 1.
The determining module 351 may be further configured to, if the number of companies is greater than 1, rank each matching score higher than the preset matching score from high to low, and determine, according to the ranking result, whether the ratio between the matching score of the first rank and the matching score of the second rank is greater than or equal to the preset ratio.
The display module 352 may be further configured to enhance and display the company corresponding to the first-ranked matching score if the first-ranked matching score is greater than or equal to the preset ratio.
The display module 352 may be further configured to add each company with a matching score higher than the preset matching score to a preset company list for display if the matching score is smaller than the preset ratio.
In a specific application scenario, the display unit includes a determining module 353, a scoring module 354, and a display module 352.
The determination module 353 may be configured to determine a target position provided by a company in the company search results.
The scoring module 354 may be configured to score the similarity total score corresponding to the target position, to obtain the scored similarity total score corresponding to the target position.
The display module 352 may be configured to sort the positions according to the score corresponding to the target position from high to low according to the added similarity total score corresponding to the target position and the similarity total scores corresponding to other positions, and determine, according to the sorting result, the company positions meeting the preset top ranking condition for display.
It should be noted that, for other corresponding descriptions of each functional module related to the information searching apparatus provided by the embodiment of the present invention, reference may be made to corresponding descriptions of the method shown in fig. 1, which are not repeated herein.
Based on the above method as shown in fig. 1, correspondingly, the embodiment of the present invention further provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, implements the following steps: acquiring a search sentence input by a user; identifying a first type entity and a second type entity contained in the search statement; obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; and displaying the company positions meeting the preset top ranking conditions according to the similarity total score.
Based on the embodiment of the method shown in fig. 1 and the apparatus shown in fig. 4, the embodiment of the present invention further provides a physical structure diagram of a computer device, as shown in fig. 6, where the computer device includes: a processor 41, a memory 42, and a computer program stored on the memory 42 and executable on the processor, wherein the memory 42 and the processor 41 are both arranged on a bus 43, the processor 41 performing the following steps when said program is executed: acquiring a search sentence input by a user; identifying a first type entity and a second type entity contained in the search statement; obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; and displaying the company positions meeting the preset top ranking conditions according to the similarity total score.
According to the technical scheme, the search statement input by the user can be obtained; identifying a first type entity and a second type entity contained in the search statement; meanwhile, a search list corresponding to the search statement is obtained, and similarity scores of the first type entity and the second type entity and field information related to company positions in the search list are calculated respectively; adding the similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position; finally, according to the similarity total score, the company positions meeting the preset ranking front condition are displayed, so that the similarity scores of the field information related to the first type entity and the second type entity and the company positions in the search sentences are calculated respectively through entity identification of the spoken or unstructured search sentences input by the user, the accurate matching of the search sentences input by the user and the different company positions can be realized, in addition, the different company positions are ordered through the similarity total score, the company positions meeting the preset ranking front condition are screened out and recommended to the user, and the company positions meeting the self requirements of the user can be recommended accurately for the user.
It will be appreciated by those skilled in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be concentrated on a single computing device, or distributed across a network of computing devices, they may alternatively be implemented in program code executable by computing devices, so that they may be stored in a memory device for execution by computing devices, and in some cases, the steps shown or described may be performed in a different order than that shown or described, or they may be separately fabricated into individual integrated circuit modules, or multiple modules or steps within them may be fabricated into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (12)

1. An information search method, comprising:
Acquiring a search sentence input by a user;
identifying a first type entity and a second type entity contained in the search statement;
obtaining a search list corresponding to the search statement, and respectively calculating similarity scores of the first type entity and the second type entity and field information related to company positions in the search list;
if the category information corresponding to the field information is a working title, merging keywords repeatedly appearing in the working title, determining word number information corresponding to the merged working title, and adjusting the similarity score according to the word number information to obtain adjusted similarity scores corresponding to the first type entity and the second type entity respectively;
if the category information corresponding to the field information is a work description, extracting keywords in the field information, and adjusting the similarity score according to the matching degree of the entity and the keywords to obtain adjusted similarity scores corresponding to the first category entity and the second category entity respectively;
if the category information corresponding to the field information is wage information, determining an industry wage range in which the wage information is located, and adjusting the similarity score according to the industry wage range to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity;
Adding the adjusted similarity scores corresponding to the first type entity and the second type entity respectively to obtain a similarity total score corresponding to the company position;
and displaying the company positions meeting the preset top ranking conditions according to the similarity total score.
2. The method of claim 1, wherein after adding the adjusted similarity scores corresponding to the first type of entity and the second type of entity, respectively, to obtain a similarity total score corresponding to the company position, the method further comprises:
calculating an industry hotness score, a company hotness score, a place hotness score and a position hotness score corresponding to the company position;
adding the industry hotness score, the company hotness score, the place hotness score and the job position hotness score to obtain a hotness total score corresponding to the company job position;
adding the heat total corresponding to the company position with the similarity total score to obtain an adjusted similarity total score corresponding to the company position;
displaying the company positions meeting the preset top ranking condition according to the similarity total score, including:
And sorting the company positions according to the adjusted similarity total score, and determining the company positions meeting the preset top ranking condition according to the sorting result to display.
3. The method of claim 1, wherein after said identifying the first type of entity and the second type of entity contained in the search statement, the method further comprises:
matching the first type entity and the second type entity with corresponding field information of each company respectively to obtain a matching result, and calculating a matching score of the search statement and each company according to the matching result;
and determining and displaying the company search results corresponding to the search sentences according to the matching scores.
4. The method of claim 3, wherein the matching the first type of entity and the second type of entity with the corresponding field information of each company respectively to obtain a matching result, and calculating the matching score of the search sentence and each company according to the matching result comprises:
the first type entity and the second type entity are respectively and strictly matched with corresponding field information of a target company in each company;
If the first type entity and/or the second type entity are/is completely matched with the corresponding field information of the target company, calculating the matching score of the search statement and the target company according to the category information corresponding to the corresponding field information;
and if the first type entity and the second type entity are not completely matched with the corresponding field information of the target company, respectively carrying out fuzzy matching on the first type entity and the second type entity with the corresponding field information of the target company to obtain fuzzy matching scores of the first type entity and the second type entity with the corresponding field information, and determining the matching scores of the search statement and the target company according to the fuzzy matching scores.
5. The method of claim 4, wherein calculating the matching score of the search term and the target company according to the category information corresponding to the corresponding field information comprises:
if a plurality of field information is matched with the first type entity and/or the second type entity, respectively calculating strict matching scores of the plurality of field information and the first type entity and/or the second type entity by adopting a preset index weighting scoring algorithm according to category information respectively corresponding to the plurality of fields;
And adding the strict matching scores corresponding to the field information to obtain the matching scores of the search statement and the target company.
6. The method of claim 4, wherein the determining a match score for the search term to the target company based on the fuzzy match score comprises:
and screening the maximum fuzzy matching score from the fuzzy matching scores, and determining the maximum fuzzy matching score as the matching score of the search sentence and the target company.
7. The method of claim 3, wherein determining and displaying company search results corresponding to the search term according to the matching score comprises:
determining the maximum matching score in the matching scores of the search sentences and the companies, and judging whether the maximum matching score is a matching score obtained through strict matching according to the score magnitude corresponding to the maximum matching score;
if the maximum matching score is a matching score obtained through strict matching, judging whether target matching scores with the same score magnitude as the maximum matching score exist in the matching scores or not;
if the target matching score which has the same score magnitude as the maximum matching score does not exist, enhancing and displaying the company corresponding to the maximum matching score;
And if the target matching score with the same score magnitude as the maximum matching score exists, adding the companies respectively corresponding to the maximum matching score and the target matching score to a preset company list for display.
8. The method of claim 7, wherein after said determining whether said maximum match score is a strict match score based on a magnitude of a score corresponding to said maximum match score, said method further comprises:
if the maximum matching score is not the matching score obtained through strict matching, determining the number of companies with the matching score higher than a preset matching score;
if the number of the companies is equal to 1, enhancing and displaying the companies with the matching scores higher than the preset matching scores;
if the number of the companies is greater than 1, sequencing all the matching scores higher than the preset matching score from high to low, and judging whether the ratio between the matching score of the first ranking and the matching score of the second ranking is greater than or equal to the preset ratio according to the sequencing result;
if the matching score is larger than or equal to the preset ratio, enhancing and displaying the company corresponding to the first matching score;
if the matching score is smaller than the preset ratio, each company with the matching score higher than the preset matching score is added to a preset company list for display.
9. A method according to claim 3, wherein said presenting company positions meeting a preset top ranking condition according to said similarity total score comprises:
determining a target position provided by a company in the company search results;
adding the similarity total score corresponding to the target position to obtain the added similarity total score corresponding to the target position;
and sorting all the positions according to the corresponding scores from high to low according to the added similarity total scores corresponding to the target positions and the similarity total scores corresponding to other positions, and determining the positions of the companies meeting the preset top ranking condition according to the sorting result to display.
10. An information search apparatus, comprising:
the acquisition unit is used for acquiring search sentences input by a user;
the identification unit is used for identifying the first type entity and the second type entity contained in the search statement;
the computing unit is used for acquiring a search list corresponding to the search statement, and respectively computing similarity scores of the first type entity and the second type entity and field information related to company positions in the search list;
The adjustment unit is used for merging the keywords repeatedly appearing in the working titles if the category information corresponding to the field information is the working titles, determining word count information corresponding to the merged working titles, and adjusting the similarity scores according to the word count information to obtain adjusted similarity scores corresponding to the first type entities and the second type entities respectively; if the category information corresponding to the field information is a work description, extracting keywords in the field information, and adjusting the similarity score according to the matching degree of the entity and the keywords to obtain adjusted similarity scores corresponding to the first category entity and the second category entity respectively; if the category information corresponding to the field information is wage information, determining an industry wage range in which the wage information is located, and adjusting the similarity score according to the industry wage range to obtain adjusted similarity scores respectively corresponding to the first type entity and the second type entity;
the adding unit is used for adding the adjusted similarity scores corresponding to the first type entity and the second type entity respectively to obtain a total score corresponding to the company position;
And the display unit is used for displaying the company positions meeting the preset ranking front condition according to the similarity total score.
11. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the computer program when executed by the processor implements the steps of the method according to any one of claims 1 to 9.
12. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 9.
CN202011576921.4A 2020-12-28 2020-12-28 Information searching method, device, storage medium and computer equipment Active CN112612961B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011576921.4A CN112612961B (en) 2020-12-28 2020-12-28 Information searching method, device, storage medium and computer equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011576921.4A CN112612961B (en) 2020-12-28 2020-12-28 Information searching method, device, storage medium and computer equipment

Publications (2)

Publication Number Publication Date
CN112612961A CN112612961A (en) 2021-04-06
CN112612961B true CN112612961B (en) 2024-02-02

Family

ID=75248438

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011576921.4A Active CN112612961B (en) 2020-12-28 2020-12-28 Information searching method, device, storage medium and computer equipment

Country Status (1)

Country Link
CN (1) CN112612961B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114490884B (en) * 2021-12-21 2023-06-06 北京三快在线科技有限公司 Method, device, electronic equipment and storage medium for determining entity association relation

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102411579A (en) * 2010-09-20 2012-04-11 腾讯科技（深圳）有限公司 Method and device for searching industry relevant information
CN103279529A (en) * 2013-05-30 2013-09-04 北京邮电大学 Unstructured data retrieval method and system
CN106447285A (en) * 2016-09-12 2017-02-22 北京大学 Multidimensional field key knowledge-based recruitment information matching method
CN107515904A (en) * 2017-07-31 2017-12-26 北京拉勾科技有限公司 A kind of position searching method and computing device
CN109558429A (en) * 2018-11-16 2019-04-02 广东百城人才网络股份有限公司 The two-way recommendation method and system of talent service based on internet big data

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10831841B2 (en) * 2016-12-15 2020-11-10 Microsoft Technology Licensing, Llc Determining similarities among job titles to enhance job searching

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102411579A (en) * 2010-09-20 2012-04-11 腾讯科技（深圳）有限公司 Method and device for searching industry relevant information
CN103279529A (en) * 2013-05-30 2013-09-04 北京邮电大学 Unstructured data retrieval method and system
CN106447285A (en) * 2016-09-12 2017-02-22 北京大学 Multidimensional field key knowledge-based recruitment information matching method
CN107515904A (en) * 2017-07-31 2017-12-26 北京拉勾科技有限公司 A kind of position searching method and computing device
CN109558429A (en) * 2018-11-16 2019-04-02 广东百城人才网络股份有限公司 The two-way recommendation method and system of talent service based on internet big data

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Akeem Olowolayemo.University Based Job Recommender &Alumni System.2018 International Conference on Computing,Engineering,and Design(ICCED).2019,全文. *
基于模糊启发式的KNN算法在人才需求信息分类中的应用;唐倩;李梁;殷志恒;;重庆理工大学学报(自然科学)(03);全文 *

Also Published As

Publication number Publication date
CN112612961A (en) 2021-04-06

Similar Documents

Publication Publication Date Title
US9009146B1 (en) Ranking search results based on similar queries
US9201880B2 (en) Processing a content item with regard to an event and a location
CN105893533B (en) Text matching method and device
US9836511B2 (en) Computer-generated sentiment-based knowledge base
JP5731250B2 (en) System and method for recommending interesting content in an information stream
CN109118336B (en) Information recommendation method and device, computer equipment and storage medium
US9405805B2 (en) Identification and ranking of news stories of interest
US7783620B1 (en) Relevancy scoring using query structure and data structure for federated search
US7617205B2 (en) Estimating confidence for query revision models
JP4809441B2 (en) Estimating search category synonyms from user logs
US7630978B2 (en) Query rewriting with spell correction suggestions using a generated set of query features
CN101622618B (en) With the search based on concept and the information retrieval system of classification, method and software
US8103650B1 (en) Generating targeted paid search campaigns
US7765209B1 (en) Indexing and retrieval of blogs
US20110119208A1 (en) Method and system for developing a classification tool
JP2009521750A (en) Analyzing content to determine context and providing relevant content based on context
US10140297B2 (en) Supplementing search results with information of interest
US20170011043A1 (en) System and method for ranking documents
CN113901308A (en) Knowledge graph-based enterprise recommendation method and recommendation device and electronic equipment
TW201617981A (en) Method and system for profiling job candidates
CN115858731A (en) Method, device and system for matching laws and regulations of law and regulation library
CN112612961B (en) Information searching method, device, storage medium and computer equipment
CN114969566A (en) Distance-measuring government affair service item collaborative filtering recommendation method
US20100057724A1 (en) Server device for creating list of general words to be excluded from search result
CN112184021B (en) Answer quality assessment method based on similar support set

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant