CN107729351A - Multilayer inquiry correcting method and system based on music searching engine - Google Patents

Multilayer inquiry correcting method and system based on music searching engine Download PDF

Info

Publication number
CN107729351A
CN107729351A CN201710757048.0A CN201710757048A CN107729351A CN 107729351 A CN107729351 A CN 107729351A CN 201710757048 A CN201710757048 A CN 201710757048A CN 107729351 A CN107729351 A CN 107729351A
Authority
CN
China
Prior art keywords
song
field
document
inquiry
index entry
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.)
Pending
Application number
CN201710757048.0A
Other languages
Chinese (zh)
Inventor
陈君炫
朱映波
吴三阳
曾荣
王伟
潘志锋
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.)
iMusic Culture and Technology Co Ltd
Original Assignee
iMusic Culture and Technology 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 iMusic Culture and Technology Co Ltd filed Critical iMusic Culture and Technology Co Ltd
Priority to CN201710757048.0A priority Critical patent/CN107729351A/en
Publication of CN107729351A publication Critical patent/CN107729351A/en
Pending legal-status Critical Current

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
    • G06F16/951—Indexing; Web crawling techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
    • G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/686—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title or artist information, time, location or usage information, user ratings

Abstract

The present invention proposes a kind of multilayer inquiry correcting method based on music searching engine, and on the premise of not having any returning result in music searching engine based on user's query word, inquiry correction is carried out to the query word of user；The present invention proposes that multilayer inquiry correct algorithm is corrected to user's query word, including user's query word is converted into phonetic progress homonym inquiry and the index entry similar to user's query word is found using editing distance algorithm, the multiple results returned after correction are as the Candidate Set for correcting inquiry；Using suitable Candidate Set screening technique, including temperature screening technique, document frequency screening technique and Longest Common Substring length screening technique, select optimal correction to inquire about in Candidate Set and substitute user's query word.The present invention can correct the malformed queries of user in search engine exactly, improve the quality of search result.

Description

Multilayer inquiry correcting method and system based on music searching engine
Technical field
The present invention relates to music data process field, is entangled more particularly to a kind of multilayer inquiry based on music searching engine Correction method and system.
Background technology
Today's society, electronic music are skyrocketed through, and people become larger to the demand for fast and effeciently retrieving music.It is near Year, with the exploration of researcher, the use of multimedia search engine realizes breakthrough growth.But index the master of song The mode of leading is still by being modeled completion to basic song attributes data.When a given inquiry, without accurate Match somebody with somebody or when the record of approximate match is to match it, music retrieval system is just difficult to find the song for meeting user inquiry.
User in a search engine input inquiry word when, common mistake is to input character by mistake, such as by " easy melt thousand Imperial or royal seal " mistake is write as " imperial or royal seals of Yi Yang thousand ".Statistical analysis, in the inquiry of search engine, about 10%~15% inquiry is mistake Input, thus how automatic and correction user exactly malformed queries, the lifting for search result quality has Very big help.
To sum up, in music searching engine, user error inquiry is very common.If it can correct user's exactly Malformed queries, it is to have very great help for the quality for improving search engine.However, in music vertical search field, at present also There is no a kind of effective and very high accuracy rate method, there is no the feelings of any returning result in music searching engine especially Under condition.
The content of the invention
Based on this, it is an object of the invention to provide a kind of multilayer inquiry correcting method based on music searching engine and it is System, on the premise of can not having any returning result in music searching engine, is efficiently corrected to the query word of user, is carried The accuracy rate of high music searching.
To reach above-mentioned purpose, the embodiment of the present invention uses following technical scheme：
A kind of multilayer inquiry correcting method based on music searching engine, including：
S1, when carrying out music searching based on user's query word, when no result returns, judge user's query word whether be Text, if performing step S2, otherwise perform step S4；
S2, user's query word is converted into phonetic, then inquires about the spelling of song key element document and the index of simplicity field , obtain required song key element document；Wherein, song key element document includes：Song key element field, song key element spelling word Section, song key element simplicity field, song element type field and song key element temperature field；
S3, obtained song key element document will be inquired about it is ranked up according to temperature, returns to the song of temperature highest document Key element field is inquired about for optimum remediation；
S4, find out in the song title field of song document and the index entry of artist name section text with user's query word This editing distance is the index entry of first threshold；Wherein, the song document include song title, Ge Shouming, album name, the lyrics with And the combined field of song title and singer's name；
S5, the index entry for determining whether matching, if performing step S6, otherwise perform step S7；
S6, the Candidate Set using the match index item of return as correction inquiry, select document frequency highest in Candidate Set Song key element field be optimum remediation inquiry；Wherein, the document frequency of the index entry refers to the document for including the index entry Quantity；
S7, to find out in the index entry of the combined field of song document with the text editing of user's query word distance be second The combined index item of threshold value；
S8, the combined index item for determining whether matching, if performing step S9, otherwise perform step S10；
S9, using the matching of return combination index entry as correction query candidate collection, select most to grow public son in Candidate Set Maximum index entry of going here and there is inquired about for optimum remediation；Wherein, the Longest Common Substring refers in the two sequences while occurred The length of most long sub-sequence；
S10, return without the prompting for correcting inquiry.
And a kind of multilayer inquiry correcting system based on music searching engine is proposed, including：
First judging unit, for when carrying out music searching based on user's query word, when no result returns, judging that user looks into Ask whether word is Chinese, if it is notify converting unit that user's query word is converted into phonetic；First query unit will inquire about The spelling of song key element document and the index entry of simplicity field, obtain required song key element document；Wherein, song key element document Including：Song key element field, song key element spelling field, song key element simplicity field, song element type field and song Key element temperature field；First corrects unit, being ranked up for will inquire about obtained song key element document according to temperature, returning to heat The song key element field for spending highest document is inquired about for optimum remediation；
Otherwise, the second query unit is notified to be looked in the song title field of song document and the index entry of artist name section Go out the index entry for first threshold with the text editing of user's query word distance；Wherein, the song document includes song title, song The combined field of hand name, album name, the lyrics and song title and singer's name；
Second judging unit, for determining whether the index entry of matching, if so, then notifying the second correction unit to return Return match index item as correct inquiry Candidate Set, in Candidate Set selection document frequency highest song key element field for Optimum remediation is inquired about；Wherein, the document frequency of the index entry refers to the number of documents for including the index entry；Otherwise the 3rd is notified It is the second threshold that query unit, which is found out in the index entry of the combined field of song document with the text editing of user's query word distance, The combined index item of value；
3rd judging unit, for determining whether the combined index item of matching, if so, then notifying the 3rd correction unit Matching combination index entry using return selects the maximum index of Longest Common Substring as query candidate collection is corrected in Candidate Set Item is inquired about for optimum remediation；Wherein, the Longest Common Substring refers to the most long sub-sequence in the two sequences while occurred Length；
Otherwise, result returning unit is notified to return without the prompting for correcting inquiry.
Using the present invention, when being returned without result, user's query word can be corrected, including user's query word is changed Homonym inquiry is carried out into phonetic and finds the index entry similar to user's query word using editing distance algorithm, is protected as far as possible User query word of the card without returning result can correct success.Using suitable Candidate Set screening technique, including temperature screening side Method, document frequency screening technique and Longest Common Substring length screening technique, select optimal correction to inquire about in Candidate Set Substitute user's query word, it is ensured that correct the accuracy of inquiry.
Brief description of the drawings
Accompanying drawing be for providing a further understanding of the present invention, and a part for constitution instruction, it is and following specific Embodiment is used to explain the present invention together, but should not be construed as limiting the invention.In the accompanying drawings,
Fig. 1 is the schematic flow sheet of the multilayer inquiry correcting method proposed by the present invention based on music searching engine；
Fig. 2 is the structural representation of the multilayer inquiry correcting system proposed by the present invention based on music searching engine.
Embodiment
For the objects, technical solutions and advantages of the present invention are more clearly understood, below in conjunction with drawings and Examples, to this Invention is described in further detail.It should be appreciated that embodiment described herein is only to explain the present invention, Do not limit protection scope of the present invention.
Fig. 1 shows a kind of schematic flow sheet of the multilayer inquiry correcting method based on music searching engine, including：
S1, judge whether user's query word is Chinese, if performing step S2, otherwise perform step S4；
Specifically, when carrying out music searching based on user's query word, when no result returns, judge user's query word whether be Chinese, if performing step S2, otherwise perform step S4.
S2, user's query word is converted into phonetic, then inquires about the spelling of song key element document and the index of simplicity field , obtain required song key element document；Wherein, song key element document includes：Song key element field, song key element spelling word Section, song key element simplicity field, song element type field and song key element temperature field；
It is song title or singer's name that song element type field, which is used for distinguishing song key element, and song key element temperature field is used To represent the temperature of song key element.Other song key element phonetic field and song key element simplicity field need to create index entry, Before creating index entry any text-processing need not be carried out to field.
S3, obtained song key element document will be inquired about it is ranked up according to temperature, returns to the song of temperature highest document Key element field is inquired about for optimum remediation；
S4, find out in the song title field of song document and the index entry of artist name section text with user's query word This editing distance is the index entry of first threshold；Wherein, the song document include song title, Ge Shouming, album name, the lyrics with And the combined field of song title and singer's name；
Song title and artist name section need to create index entry, need to be segmented before creating index entry, capital and small letter turns Change, Non-American Standard Code for Information Interchange conversion and remove the text-processing such as the possessive case, combined field need not be carried out at text before creating index Reason.
Between the editing distance refers to two word strings, as the minimum edit operation number needed for one changes into another. The edit operation of permission includes a character being substituted for another character, inserts a character, deletes a character.
Preferably, the first threshold is 1.
S5, the index entry for determining whether matching；If performing step S6, step S7 is otherwise performed；
S6, the Candidate Set using the match index item of return as correction inquiry, select document frequency highest in Candidate Set Song key element field be optimum remediation inquiry；Wherein, the document frequency of the index entry refers to the document for including the index entry Quantity；
S7, to find out in the index entry of the combined field of song document with the text editing of user's query word distance be second The combined index item of threshold value；
The combined field refers to field caused by multiple field combinations, for combined field, does not enter before creating index Any text-processing of row, the in addition different field of combined field centre space-separated.For example, song document includes song File-name field " black sweater " and artist name section " Zhou Jielun ", then combined field is exactly " black sweater Zhou Jielun " and " Zhou Jie Human relations black sweater ", index entry namely " black sweater Zhou Jielun " and " the Zhou Jielun black sweater " of generation.
Preferably, the Second Threshold is 1.
S8, the combined index item for determining whether matching；If performing step S9, step S10 is otherwise performed；
S9, using the matching of return combination index entry as correction query candidate collection, select most to grow public son in Candidate Set Maximum index entry of going here and there is inquired about for optimum remediation；Wherein, the Longest Common Substring refers in the two sequences while occurred The length of most long sub-sequence；
One subsequence occurs from the sequence of relative ranks, but is not necessarily continuous.
S10, return without the prompting for correcting inquiry.
Further, when judging the combined index item without matching in step S8, follow the steps below：
S11, to find out in the index entry of the combined field of song document with the text editing of user's query word distance be the Two threshold values add 1 combined index item；
S12, the combined index item for determining whether matching, if performing step S13, otherwise perform step S10；
S13, using the matching of return combination index entry as correction text candidates collection, select most to grow public son in Candidate Set The maximum index entry of string；
S14, judge whether the Longest Common Substring is more than threshold threshold, if performing step S15, otherwise perform step Rapid S10；
S15, the maximum index entry of Longest Common Substring is returned as optimum remediation inquiry.
The corresponding above method, the present invention propose a kind of multilayer inquiry correcting system based on music searching engine, its structure Schematic diagram such as Fig. 2, including：
First judging unit, for when carrying out music searching based on user's query word, when no result returns, judging that user looks into Ask whether word is Chinese, if it is notify converting unit that user's query word is converted into phonetic；First query unit will inquire about The spelling of song key element document and the index entry of simplicity field, obtain required song key element document；Wherein, song key element document Including：Song key element field, song key element spelling field, song key element simplicity field, song element type field and song Key element temperature field；First corrects unit, being ranked up for will inquire about obtained song key element document according to temperature, returning to heat The song key element field for spending highest document is inquired about for optimum remediation；
Otherwise, the second query unit is notified to be looked in the song title field of song document and the index entry of artist name section Go out the index entry for first threshold with the text editing of user's query word distance；Wherein, the song document includes song title, song The combined field of hand name, album name, the lyrics and song title and singer's name；
Second judging unit, for determining whether the index entry of matching, if so, then notifying the second correction unit to return Return match index item as correct inquiry Candidate Set, in Candidate Set selection document frequency highest song key element field for Optimum remediation is inquired about；Wherein, the document frequency of the index entry refers to the number of documents for including the index entry；Otherwise the 3rd is notified It is the second threshold that query unit, which is found out in the index entry of the combined field of song document with the text editing of user's query word distance, The combined index item of value；
3rd judging unit, for determining whether the combined index item of matching, if so, then notifying the 3rd correction unit Matching combination index entry using return selects the maximum index of Longest Common Substring as query candidate collection is corrected in Candidate Set Item is inquired about for optimum remediation；Wherein, the Longest Common Substring refers to the most long sub-sequence in the two sequences while occurred Length；
Otherwise, result returning unit is notified to return without the prompting for correcting inquiry.
Further, in addition to：3rd query unit, the 4th judging unit, the 5th judging unit, the 4th correct unit；
3rd query unit, for when the 3rd judging unit judge without match combined index item when, in song document Combined field index entry in find out with the text editing of user's query word distance for Second Threshold add 1 combined index item；
4th judging unit is used for the combined index item for determining whether matching, if so, then notice selecting unit will be returned The matching combination index entry returned selects the maximum index entry of Longest Common Substring as text candidates collection is corrected in Candidate Set； 5th judging unit judges whether the Longest Common Substring is more than threshold threshold, if it is, notifying the 4th correction unit to return The maximum index entry of Longest Common Substring is returned to inquire about for optimum remediation；
Otherwise, result returning unit is notified to return without the prompting for correcting inquiry.
Beneficial effects of the present invention：
1st, the present invention proposes that multilayer inquiry correct algorithm is corrected to user's query word, including user's query word is changed Homonym inquiry is carried out into phonetic and finds the index entry similar to user's query word using editing distance algorithm, is protected as far as possible User query word of the card without returning result can correct success.
2nd, using suitable Candidate Set screening technique, including temperature screening technique, document frequency screening technique and most long Public substring length screening technique, select optimal correction to inquire about in Candidate Set and substitute user's query word, it is ensured that correct inquiry Accuracy.
3rd, the present invention of the invention can exactly correct the malformed queries of user in search engine, improve search result Quality, there is very high availability.
Without departing from the thought of the invention, the various different embodiments of the present invention are combined, all should When being considered as present disclosure；In the range of the technology design of the present invention, a variety of simple modifications are carried out to technical scheme And any combination of the thought without prejudice to the invention that different embodiments are carried out, all should protection scope of the present invention it It is interior.

Claims (6)

  1. A kind of 1. multilayer inquiry correcting method based on music searching engine, it is characterised in that including：
    S1, when carrying out music searching based on user's query word, when no result returns, judge whether user's query word is Chinese, such as Fruit is to perform step S2, otherwise performs step S4；
    S2, user's query word is converted into phonetic, then inquires about the spelling of song key element document and the index entry of simplicity field, obtain To required song key element document；Wherein, song key element document includes：Song key element field, song key element spelling field, song Key element simplicity field, song element type field and song key element temperature field；
    S3, obtained song key element document will be inquired about it is ranked up according to temperature, returns to the song key element of temperature highest document Field is inquired about for optimum remediation；
    S4, find out in the song title field of song document and the index entry of artist name section and compiled with the text of user's query word Collect the index entry that distance is first threshold；Wherein, the song document includes song title, Ge Shouming, album name, the lyrics and song The combined field of song name and singer's name；
    S5, the index entry for determining whether matching, if performing step S6, otherwise perform step S7；
    S6, the Candidate Set using the match index item of return as correction inquiry, select document frequency highest song in Candidate Set Qu Yaosu fields are inquired about for optimum remediation；Wherein, the document frequency of the index entry refers to the number of documents for including the index entry；
    S7, to find out in the index entry of the combined field of song document with the text editing of user's query word distance be Second Threshold Combined index item；
    S8, the combined index item for determining whether matching, if performing step S9, otherwise perform step S10；
    S9, using the matching of return combination index entry as correcting query candidate collection, selection Longest Common Substring is most in Candidate Set Big index entry is inquired about for optimum remediation；Wherein, the Longest Common Substring refers in the two sequences while occurred most long The length of subsequence；
    S10, return without the prompting for correcting inquiry.
  2. 2. the multilayer inquiry correcting method according to claim 1 based on music searching engine, it is characterised in that work as step When the combined index item without matching is judged in S8, follow the steps below：
    S11, to find out in the index entry of the combined field of song document with the text editing of user's query word distance be the second threshold Value plus 1 combined index item；
    S12, the combined index item for determining whether matching, if performing step S13, otherwise perform step S10；
    S13, using the matching of return combination index entry as correcting text candidates collection, selection Longest Common Substring is most in Candidate Set Big index entry；
    S14, judge whether the Longest Common Substring is more than threshold threshold, if performing step S15, otherwise perform step S10；
    S15, the maximum index entry of Longest Common Substring is returned as optimum remediation inquiry.
  3. 3. the multilayer inquiry correcting method according to claim 1 based on music searching engine, it is characterised in that the song The combined field of bent document separates different field using space character.
  4. A kind of 4. multilayer inquiry correcting system based on music searching engine, it is characterised in that including：
    First judging unit, for when carrying out music searching based on user's query word, when no result returns, judging user's query word Whether it is Chinese, if it is notifies converting unit that user's query word is converted into phonetic；First query unit will inquire about song The spelling of key element document and the index entry of simplicity field, obtain required song key element document；Wherein, song key element document bag Include：Song key element field, song key element spelling field, song key element simplicity field, song element type field and song will Plain temperature field；First corrects unit, being ranked up for will inquire about obtained song key element document according to temperature, returning to temperature The song key element field of highest document is inquired about for optimum remediation；Otherwise, notify the second query unit in the song of song document The index for first threshold with the text editing of user's query word distance is found out in the index entry of file-name field and artist name section ；Wherein, the song document includes the combined characters of song title, Ge Shouming, album name, the lyrics and song title and singer's name Section；
    Second judging unit, for determining whether the index entry of matching, if so, then notifying the second correction unit by return For match index item as the Candidate Set for correcting inquiry, it is optimal that document frequency highest song key element field is selected in Candidate Set Correct inquiry；Wherein, the document frequency of the index entry refers to the number of documents for including the index entry；Otherwise the 3rd inquiry is notified It is Second Threshold that unit, which is found out in the index entry of the combined field of song document with the text editing of user's query word distance, Combined index item；
    3rd judging unit, for determining whether the combined index item of matching, if so, then notifying the 3rd correction unit to return The matching combination index entry returned as query candidate collection is corrected, selected in Candidate Set the maximum index entry of Longest Common Substring for Optimum remediation is inquired about；Wherein, the Longest Common Substring refers to the length of the most long sub-sequence in the two sequences while occurred； Otherwise, result returning unit is notified to return without the prompting for correcting inquiry.
  5. 5. the multilayer inquiry correcting system according to claim 4 based on music searching engine, it is characterised in that also wrap Include：3rd query unit, the 4th judging unit, the 5th judging unit, the 4th correct unit；
    3rd query unit, for when the 3rd judging unit judge without match combined index item when, in the group of song document Close and found out in the index entry of field with the text editing of user's query word apart from the combined index item for adding 1 for Second Threshold；
    4th judging unit is used for the combined index item for determining whether matching, if so, then notifying selecting unit by return Matching combination index entry selects the maximum index entry of Longest Common Substring as text candidates collection is corrected in Candidate Set；5th Judging unit judges whether the Longest Common Substring is more than threshold threshold, if it is, notifying the 4th correction unit to return most The maximum index entry of long public substring is inquired about for optimum remediation；
    Otherwise, result returning unit is notified to return without the prompting for correcting inquiry.
  6. 6. the multilayer inquiry correcting system according to claim 4 based on music searching engine, it is characterised in that the song The combined field of bent document separates different field using space character.
CN201710757048.0A 2017-08-29 2017-08-29 Multilayer inquiry correcting method and system based on music searching engine Pending CN107729351A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710757048.0A CN107729351A (en) 2017-08-29 2017-08-29 Multilayer inquiry correcting method and system based on music searching engine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710757048.0A CN107729351A (en) 2017-08-29 2017-08-29 Multilayer inquiry correcting method and system based on music searching engine

Publications (1)

Publication Number Publication Date
CN107729351A true CN107729351A (en) 2018-02-23

Family

ID=61205395

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710757048.0A Pending CN107729351A (en) 2017-08-29 2017-08-29 Multilayer inquiry correcting method and system based on music searching engine

Country Status (1)

Country Link
CN (1) CN107729351A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113553398A (en) * 2021-07-15 2021-10-26 杭州网易云音乐科技有限公司 Search word correcting method and device, electronic equipment and computer storage medium

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101206672A (en) * 2007-12-25 2008-06-25 北京科文书业信息技术有限公司 Commercial articles searching non result intelligent processing system and method
CN101241514A (en) * 2008-03-21 2008-08-13 北京搜狗科技发展有限公司 Method for creating error-correcting database, automatic error correcting method and system
CN101751430A (en) * 2008-12-12 2010-06-23 汉王科技股份有限公司 Electronic dictionary fuzzy searching method
CN103116607A (en) * 2013-01-18 2013-05-22 中国传媒大学 Full-text retrieval method based on pinyin
CN103198149A (en) * 2013-04-23 2013-07-10 中国科学院计算技术研究所 Method and system for query error correction
CN103336850A (en) * 2013-07-24 2013-10-02 昆明理工大学 Method and device for confirming index word in database retrieval system
CN105206274A (en) * 2015-10-30 2015-12-30 北京奇艺世纪科技有限公司 Voice recognition post-processing method and device as well as voice recognition system

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101206672A (en) * 2007-12-25 2008-06-25 北京科文书业信息技术有限公司 Commercial articles searching non result intelligent processing system and method
CN101241514A (en) * 2008-03-21 2008-08-13 北京搜狗科技发展有限公司 Method for creating error-correcting database, automatic error correcting method and system
CN101751430A (en) * 2008-12-12 2010-06-23 汉王科技股份有限公司 Electronic dictionary fuzzy searching method
CN103116607A (en) * 2013-01-18 2013-05-22 中国传媒大学 Full-text retrieval method based on pinyin
CN103198149A (en) * 2013-04-23 2013-07-10 中国科学院计算技术研究所 Method and system for query error correction
CN103336850A (en) * 2013-07-24 2013-10-02 昆明理工大学 Method and device for confirming index word in database retrieval system
CN105206274A (en) * 2015-10-30 2015-12-30 北京奇艺世纪科技有限公司 Voice recognition post-processing method and device as well as voice recognition system

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113553398A (en) * 2021-07-15 2021-10-26 杭州网易云音乐科技有限公司 Search word correcting method and device, electronic equipment and computer storage medium
CN113553398B (en) * 2021-07-15 2024-01-26 杭州网易云音乐科技有限公司 Search word correction method, search word correction device, electronic equipment and computer storage medium

Similar Documents

Publication Publication Date Title
KR101190230B1 (en) Phrase identification in an information retrieval system
Constantin et al. PDFX: fully-automated PDF-to-XML conversion of scientific literature
Lu et al. Annotating structured data of the deep Web
CN100478949C (en) Query rewriting with entity detection
KR101223173B1 (en) Phrase-based indexing in an information retrieval system
CN103631948B (en) Identifying method of named entities
US7890859B2 (en) Rapid similarity links computation for table of contents determination
US20020021838A1 (en) Adaptively weighted, partitioned context edit distance string matching
US7483903B2 (en) Unsupervised learning tool for feature correction
KR100903961B1 (en) Indexing And Searching Method For High-Demensional Data Using Signature File And The System Thereof
CN102902826B (en) A kind of image method for quickly retrieving based on reference picture index
CN104537116A (en) Book search method based on tag
CN104991905A (en) Method for mathematical expression retrieval based on hierarchical indexing
CN109902142B (en) Character string fuzzy matching and query method based on edit distance
CN102081660B (en) Method for searching and sequencing keywords of XML documents based on semantic correlation
CN106708929A (en) Video program search method and device
CN115563313A (en) Knowledge graph-based document book semantic retrieval system
CN109062876B (en) A kind of similar web page lookup method and system based on DOM webpage beta pruning
CN102915381A (en) Multi-dimensional semantic based visualized network retrieval rendering system and rendering control method
CN107729351A (en) Multilayer inquiry correcting method and system based on music searching engine
CN105574004A (en) Webpage deduplication method and device
JP2001184358A (en) Device and method for retrieving information with category factor and program recording medium therefor
JP4826622B2 (en) Document search apparatus, search method and program
CN106126618A (en) Email address based on name recommends method and system
Packer et al. Cost effective ontology population with data from lists in ocred historical documents

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20180223

RJ01 Rejection of invention patent application after publication