CN105653661A - Search result re-ranking method and device - Google Patents
Search result re-ranking method and device Download PDFInfo
- Publication number
- CN105653661A CN105653661A CN201511008470.3A CN201511008470A CN105653661A CN 105653661 A CN105653661 A CN 105653661A CN 201511008470 A CN201511008470 A CN 201511008470A CN 105653661 A CN105653661 A CN 105653661A
- Authority
- CN
- China
- Prior art keywords
- industry
- retrieval
- ranking value
- retrieval result
- value
- 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
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
- G06F16/903—Querying
- G06F16/9038—Presentation of query results
Abstract
Embodiments of the invention disclose a search result re-ranking method and device. The method comprises the following steps: obtaining search results and calculating a time ranking value according to time information in the search results; carrying out statistics on access frequency of the search results and calculating an access frequency ranking value; calculating quantity of industry words in the search results according to a preset target word bank which comprises one or more types of industry words in electric power, spaceflight, energy and medical science; calculating a position ranking value according to the original ranking position of the search results; weighting to calculate the ranking values to obtain a total ranking value; and carrying out re-ranking on the search results according to the total ranking value. According to the method, the influences, caused by the time, access frequency, search industry relevancy and the original ranking positions in the search results, on the research results are comprehensively considered, and the search results which have high search industry relevancy and new time and which are mostly related to the user query contents are ranked foremost, so that the search quality is effectively improved, the user query time is shortened and high search efficiency is provided.
Description
Technical field
The present invention relates to technical field of information retrieval, particularly relate to a kind of retrieval result rearrangement method and device.
Background technology
Power grid enterprises are when carrying out technological innovation, it is necessary to carry out the collection of the technical information such as new technique, new method, and carry out creationary innovation and expansion based on the technical information collected. Based on search engine, user is after input inquiry content, and the retrieval result comprising the information such as title, summary, time, link can be supplied to user according to inquiry content and consult by search engine.
Currently in order to facilitate user to consult, generally described retrieval result being ranked up, the method for sequence is mainly ranked up according to matching degree, visit capacity, page quality etc., and with clooating sequence, described retrieval result is showed user. But, in actual use, such as electric power network technique personnel input inquiry content is the technical information that " transformator " is relevant to obtain transformator, and the information of ranking former is often the news contents such as fire for transformer in retrieval result, uncorrelated with transformer technology, retrieval quality is poor, and the time that user needs cost extra could inquire effective technical information in retrieval result, has a strong impact on recall precision.
Summary of the invention
The embodiment of the present invention provides a kind of retrieval result rearrangement method and device, to solve the problem that retrieval outcome quality is poor, recall precision is low of the prior art.
In order to solve above-mentioned technical problem, the embodiment of the invention discloses following technical scheme:
The embodiment of the invention discloses a kind of retrieval result rearrangement method, the method includes:
Obtain retrieval result;
According to the temporal information in retrieval result, calculate every time-sequencing value corresponding to retrieval result;
The access times of statistics retrieval result;
According to described access times, calculate every access times ranking value corresponding to retrieval result;
According to default target dictionary, calculate in retrieval result, the quantity that in described target dictionary, vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science;
According to the quantity that vocabulary in described target dictionary occurs, calculate every retrieval industry relevancy ranking value corresponding to retrieval result;
Utilize the former sorting position of retrieval result, calculate name placement value;
Described in weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value;
According to described total ranking value, described retrieval result is reset.
Preferably, described according to the temporal information in retrieval result, calculate every time-sequencing value corresponding to retrieval result, including:
Obtaining reference time information, described reference time information includes time and month;
Calculate the time difference of every temporal information corresponding to retrieval result and described reference time information;
According to described time difference, calculate and obtain described time-sequencing value.
Preferably, the target dictionary that described basis is preset, calculate in retrieval result, the quantity that in described target dictionary, vocabulary occurs, including:
Obtain category of employment information, and according to described category of employment information, from described target dictionary, select the industry vocabulary of corresponding category of employment;
The title in result and described industry vocabulary are retrieved in comparison, calculate heading remittance number;
The summary in result and described industry vocabulary are retrieved in comparison, calculate summary vocabulary number.
Preferably, the described quantity occurred according to vocabulary in described target dictionary, calculate every retrieval industry relevancy ranking value corresponding to retrieval result, including:
Preset title coefficient;
Utilize described title coefficient, described heading remittance number and described summary vocabulary number, calculate the retrieval industry degree of association obtaining retrieval result;
According to described retrieval industry degree of association, calculate and obtain retrieval industry relevancy ranking value.
Preferably, described in described weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value, including:
Preset weight coefficient, and described weight coefficient includes time coefficient, access times coefficient, retrieval industry correlation coefficient and position parameter;
According to the temporal information in described time coefficient and retrieval result, calculate weight temporal ranking value; According to described access times coefficient and access times, calculate weighting access times ranking value; According to the quantity that vocabulary in described retrieval industry correlation coefficient and target dictionary occurs, calculate Weighted Searching industry relevancy ranking value; Former sorting position according to described position parameter and retrieval result, calculates weighting position ranking value;
Cumulative described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and described weighting position ranking value, it is thus achieved that described total ranking value.
The embodiment of the invention also discloses a kind of retrieval result rearrangement device, this device includes:
Retrieval result acquisition module, is used for obtaining retrieval result;
Time-sequencing value computing module, for according to the temporal information in retrieval result, calculating every time-sequencing value corresponding to retrieval result;
Access times statistical module, for adding up the access times of retrieval result;
Access times ranking value computing module, for according to described access times, calculating every access times ranking value corresponding to retrieval result;
Vocabulary quantity computing module, for according to the target dictionary preset, calculating in retrieval result, the quantity that in described target dictionary, vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science;
Retrieval industry relevancy ranking value computing module, for the quantity occurred according to vocabulary in described target dictionary, calculates every retrieval industry relevancy ranking value corresponding to retrieval result;
Name placement value computing module, for utilizing the former sorting position of retrieval result, calculates name placement value;
Total ranking value computing module, described in weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value;
Order module, for according to described total ranking value, resetting described retrieval result.
Preferably, described time-sequencing value computing module also includes:
Reference time acquisition module, is used for obtaining reference time information, and described reference time information includes time and month;
Time difference computing module, for according to the temporal information in retrieval result and described reference time information, calculating the time difference of every temporal information corresponding to retrieval result and reference time information;
Time-sequencing value obtains module, for according to described time difference, calculating and obtain described time-sequencing value.
Preferably, described vocabulary quantity computing module includes:
Industry bilingual lexicon acquisition module, is used for obtaining category of employment information, and according to described category of employment information, selects the industry vocabulary of corresponding category of employment from described target dictionary;
Title vocabulary gauge calculates module, retrieves the title in result and described industry vocabulary for comparison, calculates heading remittance number;
Summary vocabulary quantity computing module, retrieves the summary in result and described industry vocabulary for comparison, calculates summary vocabulary number.
Preferably, described retrieval industry relevancy ranking value computing module includes:
Title coefficient presetting module, is used for presetting title coefficient;
Retrieval industry degree of association obtains module, is used for utilizing described title coefficient, described heading remittance number and described summary vocabulary number, calculates the retrieval industry degree of association obtaining retrieval result;
Retrieval industry relevancy ranking value obtains module, for according to described retrieval industry degree of association, calculating and obtain retrieval industry relevancy ranking value.
Preferably, described total ranking value computing module includes:
Weight coefficient presetting module, is used for presetting weight coefficient, and described weight coefficient includes time coefficient, access times coefficient, retrieval industry correlation coefficient and predeterminated position coefficient;
Weighting ranking value computing module, for according to the temporal information in described time coefficient and retrieval result, calculating weight temporal ranking value; According to described access times coefficient and access times, calculate weighting access times ranking value; According to the quantity that vocabulary in described retrieval industry correlation coefficient and target dictionary occurs, calculate Weighted Searching industry relevancy ranking value; Former sorting position according to described position parameter and retrieval result, calculates weighting position ranking value;
Accumulator module, is used for add up described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and described weighting position ranking value, it is thus achieved that described total ranking value.
From above technical scheme, the retrieval result rearrangement method of embodiment of the present invention offer and device, by obtaining retrieval result, according to the temporal information in retrieval result, calculate time-sequencing value; The access times of statistics retrieval result, and calculate access times ranking value; According to the default target dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science that includes, calculate the quantity that industry vocabulary described in retrieval result occurs, and then calculate retrieval industry relevancy ranking value; Then utilize the former sorting position of retrieval result, calculate name placement value; The above-mentioned ranking value of last weighted calculation, obtains total ranking value, sorts according to described total ranking value. Said process has considered the time in retrieval result, access times, retrieval industry degree of association and the impact on retrieval result of the former sorting position, by high for retrieval industry degree of association, the time is new, user inquires about the maximally related retrieval result of content and comes foremost, it is effectively improved retrieval quality, reduce user's query time, there is significantly high recall precision.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, the accompanying drawing used required in embodiment or description of the prior art will be briefly described below, apparently, for those of ordinary skills, under the premise not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
A kind of schematic flow sheet retrieving result rearrangement method that Fig. 1 provides for the embodiment of the present invention;
The schematic flow sheet of a kind of time-sequencing value calculating method that Fig. 2 provides for the embodiment of the present invention;
The schematic flow sheet of a kind of vocabulary quantity computational methods that Fig. 3 provides for the embodiment of the present invention;
A kind of schematic flow sheet retrieving industry relevancy ranking value calculating method that Fig. 4 provides for the embodiment of the present invention;
The schematic flow sheet of a kind of total ranking value computational methods that Fig. 5 provides for the embodiment of the present invention;
A kind of structural representation retrieving result rearrangement device that Fig. 6 provides for the embodiment of the present invention;
The structural representation of a kind of time-sequencing value computing module that Fig. 7 provides for the embodiment of the present invention;
The structural representation of a kind of vocabulary quantity computing module that Fig. 8 provides for the embodiment of the present invention;
A kind of structural representation retrieving industry relevancy ranking value computing module that Fig. 9 provides for the embodiment of the present invention;
The structural representation of a kind of total ranking value computing module that Figure 10 embodiment of the present invention provides.
Detailed description of the invention
In order to make those skilled in the art be more fully understood that the technical scheme in the present invention, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only a part of embodiment of the present invention, rather than whole embodiments. Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under not making creative work premise, all should belong to the scope of protection of the invention.
Technological innovation is broadly divided into Three models: autonomous innovation, initiating creativity and cooperative innovation. At present, power grid enterprises' technological innovation with initiating creativity be chief commander's new technique, new method combines with current electric grid production practices. Initiating creativity refers to that Innovation Main Body passes through legal means introduction innovation achievement a kind of innovation form improved on this basis under the demonstration impact and interests induction of one-up innovation.In initiating creativity process, the combination of forward position new technique, new method collection and electrical network production practices thereof can be abstract in natural mode; New technique, new method collection process in, user is generally by search engine, input inquiry content retrieves result accordingly to obtain in a search engine, described retrieval result will be shown to user with the described maximally related content priority of inquiry content, facilitate user to consult, improve recall precision.
Referring to Fig. 1, for a kind of schematic flow sheet retrieving result rearrangement method that the embodiment of the present invention provides, described sort method comprises the following steps:
Step S101: obtain retrieval result.
Described retrieval result is the inquiry content according to user, the retrieval result of return; Described retrieval result is generally returned by retrieval server, consults for user. In described retrieval result, the information such as including title, summary, time, link, and showing in a browser with certain sequence of positions with a plurality of retrieval result. As shown in Table 1, the one retrieval result provided for the embodiment of the present invention. In embodiments of the present invention, results, every all corresponding corresponding former sorting position of described retrieval result, time, title and summary are retrieved including 5. Certainly, in the specific implementation, the retrieval result shown in table one is only the present embodiment exemplary results, can also include other information such as chained address, author, Business Name in described retrieval result.
Table one:
| Former sorting position | Time | Title | Summary |
| 1 | 2013-4-12 | First title | First summary |
| 2 | 2015-9-18 | Second title | Second summary |
| 3 | 2014-5-26 | 3rd title | 3rd summary |
| 4 | 2015-2-04 | 4th title | 4th summary |
| 5 | 2014-11-21 | 5th title | 5th summary |
Step S102: according to the temporal information in retrieval result, calculates every time-sequencing value corresponding to retrieval result.
By the above-mentioned steps S101 retrieval result obtained, the temporal information that in described retrieval result, the first retrieval result is corresponding is 2013-4-12, i.e. on April 12nd, 2013; Second temporal information retrieving result corresponding is 2015-9-18, and the 3rd temporal information retrieving result corresponding is 2014-5-26, and the 4th temporal information retrieving result corresponding is 2015-2-04, and the 5th temporal information retrieving result corresponding is 2014-11-21. Described temporal information is it is to be understood that time of creating of described retrieval result or time of amendment.
Referring to the schematic flow sheet of Fig. 2 a kind of time-sequencing value calculating method provided for the embodiment of the present invention, this time-sequencing value calculating method comprises the following steps:
Step S1021: obtaining reference time information, described reference time information includes time and month.
Described reference time information is user's reference time information set in advance, described reference time information can be the random time before the current time of user search or current time, such as user was on October 20th, 2015, input inquiry content carries out technology retrieval, it is determined that described reference time information is 2015-10-20.
Step S1022: calculate the time difference of every temporal information corresponding to retrieval result and described reference time information.
In the specific implementation, described time difference is month difference, and namely the month between retrieval result temporal information and reference time information is poor, certainly in the specific implementation, described time difference can also be time difference or day difference etc.
Described month, the computing formula of difference was as follows:
Month difference=(reference time time-time retrieval result time) 12+ (reference time month-month retrieval result time)
Thus may determine that the first retrieval result corresponding month difference for (2015-2013) 12+ (10-4)=30, the second month difference retrieving result corresponding is 1, the 3rd month difference retrieving result corresponding is 17, the 4th month difference retrieving result corresponding is 8, and the 5th month difference retrieving result corresponding is 11.
Step S1023: according to described time difference, calculates and obtains described time-sequencing value.
In embodiments of the present invention, described time-sequencing value calculates in the following manner:
Time-sequencing value=((time difference+1) 1) ^ (-1)
By above-mentioned formula, it is possible to the time-sequencing value calculating the first retrieval result corresponding is ((30+1) 1)^(-1)=0.032, the equally possible time-sequencing value 0.111 and the 5th correspondingly calculating time-sequencing value the 0.056, the 4th retrieval result corresponding to time-sequencing value the 0.5, the 3rd retrieval result corresponding to described second retrieval result corresponding retrieves the time-sequencing value 0.083 that result is corresponding. By calculating every time-sequencing value corresponding to retrieval result, the reference time described in the more big time gap characterizing retrieval result of described time-sequencing value is more near, from time angle, it is ensured that newer retrieval result is aligned to position above.
Step S103: the access times of statistics retrieval result.
After user obtains retrieval result from search server, it is linked into, by what click retrieval result, the technical information that corresponding page interrogation is concrete, by the historical record that statistic record user accesses, counts the corresponding access times retrieving result. The information such as such as the first retrieval result uniquely corresponding corresponding temporal information, the first title, the second summary and the first chained address, by browser or the historical record extracting user from the temporary file that system preserves, contrast described information and historical record, judge that before and after the first retrieval result, 20 time points are accessed by the user, then the access times adding up described first retrieval result are 20 times. The equally possible access times determining other retrieval results in embodiments of the present invention, the such as access times of the second retrieval result are 7 times, the access times of the 3rd retrieval result are 13 times, and the access times of the 4th retrieval result are the access times of 30 times and the 5th retrieval result is 17 inferior.
Step S104: according to described access times, calculates every access times ranking value corresponding to retrieval result.
Described access times ranking value calculates according to below equation:
Access times ranking value=1-((access times+1) 1) ^ (-1)
Thereby determine that access times ranking value corresponding to the first retrieval result is 1-((20+1) 1) ^ (-1)=0.952, can calculate respectively also according to above-mentioned computing formula and obtain access times ranking value the 0.875, the 3rd corresponding to the second retrieval result and retrieve access times ranking value the 0.929, the 4th corresponding to result and retrieve access times ranking value corresponding to result 0.968 and access times ranking value 0.944 corresponding to the 5th retrieval result.
When actual retrieval, access times are more high show this retrieval result and user need the technical information of retrieval closer to, by calculating described access times ranking value, access times ranking value is more high shows that the number of times retrieving result accessed by the user is more many, and then ensure retrieval results many for access times is come forward position, facilitate user to consult.
Step S105: according to default target dictionary, calculates in retrieval result, the quantity that in described target dictionary, industry vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science.
In inventive embodiments, if user is the technical staff of power system, then can set that described target dictionary is electric power dictionary, described target dictionary includes power industry vocabulary;Certainly, in order to improve versatility, described target dictionary can include the industry vocabulary of multiple classification simultaneously, for instance include the industry vocabulary of electric power, space flight, the energy and medical domain simultaneously. The type of organization of described target dictionary does not limit in embodiments of the present invention, such as described target dictionary can be a dictionary that include all above-mentioned fields industry vocabulary, that be separated with profession identity, or described target dictionary can include the sub-dictionary of multiple industries such as electron dictionary, the sub-dictionary of space flight.
Referring to Fig. 3, for the schematic flow sheet of a kind of vocabulary quantity computational methods that the embodiment of the present invention provides, described vocabulary quantity computational methods comprise the following steps:
Step S1051: obtain the category of employment information of target dictionary, and according to described category of employment information, select the industry vocabulary of corresponding category of employment from described target dictionary.
In order to improve degree of freedom, user can customize the category of employment of target dictionary, consider following application scenarios: user has only to the technical information of retrieval power domain, then can set that the category of employment information of described target dictionary is electric power, and select the power industry vocabulary in described target dictionary or include the sub-dictionary of power industry of power industry vocabulary, the basis of industry degree of association is retrieved as subsequent calculations power industry; If user needs the technical information of retrieval other field, such as retrieval energy industry field, then can set that the category of employment information of described target dictionary is the energy, and then select the energy industry vocabulary in target dictionary accordingly, or including the sub-dictionary of energy industry of energy industry vocabulary, retrieve the basis of industry degree of association as subsequent calculations energy industry.
Step S1052: the industry vocabulary that the title in result is corresponding with described category of employment information is retrieved in comparison, calculates heading remittance number.
In the specific implementation, extract first retrieval result in the first title, described first title is text header section, described text header section is carried out participle, filter meaningless word process obtain title vocabulary; According to the step S1051 category of employment information set, for instance described category of employment information is electric power, then whether title vocabulary described in contrast judgement occurs in power industry vocabulary, if there is, then add up the number of times of appearance, thus obtaining heading remittance number. Pass through said process, it is determined that the first heading remittance number retrieving result corresponding is 2, namely occurs in that 2 power industry vocabulary in the title of described first retrieval result; Equally, it is determined that the second heading remittance number retrieving result corresponding is 1, the 3rd heading remittance number retrieving result corresponding is 0, and the 4th heading remittance number retrieving result corresponding is 3, and the 5th heading remittance number retrieving result corresponding is 1.
Step S1053: the industry vocabulary corresponding with described industry analogy information of making a summary in comparison retrieval result, calculates summary vocabulary number.
Similar with step S1052, same extract the first summary that the first retrieval result is corresponding, described first summary is text snippet section, described text snippet section is carried out participle, filters meaningless word and process and obtain summary vocabulary; If the category of employment information that step S1051 sets is as electric power, then whether vocabulary of making a summary described in contrast judgement occurs in power industry vocabulary, if there is, then add up the number of times of appearance, thus obtaining summary vocabulary number. Pass through said process, it is determined that the first summary vocabulary number retrieving result corresponding is 5, namely the summary of described first retrieval result occurs in that 5 power industry vocabulary;Equally, it is determined that the second summary vocabulary number retrieving result corresponding is 7, the 3rd summary vocabulary number retrieving result corresponding is 13, and the 4th summary vocabulary number retrieving result corresponding is 30, and the 5th heading remittance number retrieving result corresponding is 17.
Heading remittance number and the summary vocabulary number determined by above-mentioned steps are characterized, the quantity that the industry vocabulary of target dictionary occurs in retrieval result. Certainly, in the specific implementation, when computing hardware allows, the quantity that the vocabulary of described target dictionary occurs can also be undertaken by the mode of full-text search, and is not limited solely in title and summary.
Step S106: the quantity occurred according to industry vocabulary in described target dictionary, calculates every retrieval industry relevancy ranking value corresponding to retrieval result.
By described retrieval industry relevancy ranking value, it is ensured that retrieval result needs the retrieval result that the industry of retrieval retrieves industry degree of association higher come forward position with user, thus being further ensured that recall precision.
Referring to Fig. 4, for a kind of schematic flow sheet retrieving industry relevancy ranking value calculating method that the embodiment of the present invention provides, the method comprises the following steps:
Step S1061: preset title coefficient.
Retrieval result includes title and summary info, and wherein heading message describes the technical characteristic of retrieval result more accurately, is therefore increased the weight of title division by described title coefficient. In the specific implementation, described title coefficient is set to 3. Certainly, according to actual retrieval situation, it is arbitrary value that those skilled in the art can arrange described title coefficient, for instance arranging described title coefficient is 2 or 3.3 etc.
Step S1062: utilize described title coefficient, described heading remittance number and described summary vocabulary number, calculate the retrieval industry degree of association obtaining retrieval result.
The computing formula of described retrieval industry degree of association is as follows:
Retrieval industry degree of association=heading remittance number title coefficient+summary vocabulary number
By above-mentioned formula, calculating and obtaining the first retrieval retrieval industry degree of association corresponding to result is 2 3+5=11, is similarly obtained retrieval industry degree of association corresponding to retrieval industry degree of association corresponding to the second retrieval result, the 3rd retrieval result, the 4th retrieves retrieval industry degree of association corresponding to result and the 5th and retrieve the retrieval industry degree of association that result is corresponding.
Step S1063: according to described retrieval industry degree of association, calculates and obtains retrieval industry relevancy ranking value.
The computing formula of described retrieval industry relevancy ranking value is as follows:
Retrieval industry relevancy ranking value=1-((retrieval industry degree of association+1) 1) ^ (-1)
Thus, calculate and determine that retrieval industry relevancy ranking value corresponding to the first retrieval result is 1-((11+1) 1) ^ (-1)=0.917, it is similarly obtained the retrieval industry relevancy ranking value 0.909 that the second retrieval result is corresponding, the retrieval industry relevancy ranking value 0.667 that 3rd retrieval result is corresponding, the 4th retrieves retrieval industry relevancy ranking value corresponding to result 0.929 and the 5th retrieves the retrieval industry relevancy ranking value 0.857 that result is corresponding.
Step S107: utilize the former sorting position of retrieval result, calculates name placement value.
The retrieval result that search server returns, rule is ranked up in a certain order, for instance according to the matching degree order sequence etc. with inquiry content, the retrieval result therefore obtained is respectively provided with former sorting position.In embodiments of the present invention, the described first former sorting position retrieving result corresponding is 1, namely retrieves result from search server by described first and makes number one; Retrieve result similarly for second and retrieve result to the 5th, corresponding former sorting position 2-5 respectively.
Described name placement value is calculated by following formula:
Name placement value=((former sorting position+1) 1) ^ (-1)
Thus, the first name placement value retrieving result corresponding is ((1+1) 1) ^ (-1)=0.5, it is 0.333 that name placement value corresponding to result is retrieved in equally possible calculating second, the 3rd name placement value retrieving result corresponding is 0.25, the 4th name placement value retrieving result corresponding is 0.2, and the 5th name placement value retrieving result corresponding is 0.167.
The former sorting position described name placement value of calculating according to the retrieval result that search server returns, owing to former sorting position can be higher with the matching degree of inquiry content, reset in influence factor it is thus desirable to former sorting position to be added retrieval result, be further ensured that the correctness of ranking results.
Step S108: time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value described in weighted calculation, obtains total ranking value.
Consider the retrieval time of result, access times, industry retrieval industry degree of association and former sorting position four aspect factor, so that it is determined that final retrieval result is reset. In the specific implementation, cumulative described time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, thus calculating every total ranking value corresponding to retrieval result. As shown in Table 2, the total ranking value result provided for the embodiment of the present invention. It is calculated as example with the first total ranking value retrieving result, total ranking value=the first retrieval time-sequencing value the 0.032+ first corresponding to result of described first retrieval result retrieves access times ranking value 0.952+ corresponding to the result first retrieval industry relevancy ranking value 0.917+ first retrieving result corresponding and retrieves the name placement value 0.500 that result is corresponding, thus obtaining total ranking value corresponding to the first retrieval result is 2.401. Also according to above-mentioned computational methods, can calculate total ranking value corresponding to the second retrieval result is 2.671,3rd retrieval total ranking value corresponding to result is 1.901, total ranking value corresponding to the 4th retrieval result be 2.207 and the 5th total ranking value corresponding to retrieval result be 2.052.
Table two:
In actual applications, the Search Requirement according to user, the influence degree that retrieval result is finally sorted by time, access times, retrieval industry degree of association and former sorting position respectively can be different; Such as user considers to need the retrieval result that preferentially the acquisition time is nearest, then time factor is to pay the utmost attention to factor, then the weight of corresponding time-sequencing value needs to strengthen; User needs to get rid of as far as possible user and loses interest in the retrieval result of industry, then retrieval industry degree of association factor is to pay the utmost attention to factor, then the weight of corresponding retrieval industry relevancy ranking value needs increasing. Therefore to meet user's flexible configuration to different affecting factors weight, Optimal scheduling strategy, the embodiment of the present invention additionally provides a kind of total ranking value computational methods, referring to Fig. 5, for the schematic flow sheet of a kind of total ranking value computational methods that the embodiment of the present invention provides, described computational methods comprise the following steps:
Step S1081: preset weight coefficient, and described weight coefficient includes time coefficient, access times coefficient, retrieval industry correlation coefficient and position parameter.
Step S1082: calculate weight temporal ranking value, weighting access times ranking value, Weighted Searching industry relevancy ranking value and weighting position ranking value.
In above-mentioned steps, the embodiment of the present invention has calculated time-sequencing value, access times ranking value, Weighted Searching industry relevancy ranking value and weighting position ranking value, and above-mentioned ranking value can be understood as the ranking value that weight coefficient is 1. According to the step S1081 weight coefficient including time coefficient, access times coefficient, retrieval industry correlation coefficient and position parameter determined, calculate weight temporal ranking value, weighting access times ranking value, Weighted Searching industry relevancy ranking value and weighting position ranking value respectively.
For the calculating of weight temporal ranking value, introducing described time coefficient in the computing formula that step S1023 is corresponding, the computing formula of described weight temporal ranking value is:
Weight temporal ranking value=((time difference+1) time coefficient) ^ (-1)
And then, calculate described weight temporal ranking value according to time difference and described time coefficient.
For the calculating of weighting access times ranking value, introducing described access times coefficient in the computing formula that step S104 is corresponding, the computing formula obtaining weighting access times ranking value is:
Weighting access times ranking value=1-((access times+1) access times coefficient) ^ (-1)
Thus, weighting access times ranking value according to described access times and described access times coefficient calculations.
For the calculating of Weighted Searching industry relevancy ranking value, introducing described retrieval industry correlation coefficient in the computing formula that step S1063 is corresponding, obtaining Weighted Searching industry relevancy ranking value computing formula is:
Weighted Searching industry relevancy ranking value=1-((retrieval industry degree of association+1) retrieval industry correlation coefficient) ^ (-1)
And then by described retrieval industry degree of association and described retrieval industry correlation coefficient, calculate and obtain described Weighted Searching industry relevancy ranking value.
For the calculating of weighting position ranking value, introducing described position parameter in the computing formula that step S107 is corresponding, obtaining weighting position ranking value computing formula is:
Weighting position ranking value=((former sorting position+1) position parameter) ^ (-1)
By described former sorting position and described position parameter, calculate and determine described weighting position ranking value.
Step S1083: cumulative described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and described weighting position ranking value, it is thus achieved that described total ranking value.
According to step S1082 described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and the described weighting position ranking value obtained, the ranking value after above-mentioned weighting is added up and obtains every total ranking value corresponding to retrieval result.
Step S109: according to described total ranking value, resets described retrieval result.
According to the total ranking value determined in step S108, again retrieval result is arranged according to the order from big to small of described total ranking value. As shown in Table 3, for the retrieval result after the rearrangement of embodiment of the present invention offer. Due to second retrieve result total ranking value 2.617 > first retrieval result total ranking value 2.401 > the 4th retrieval result total ranking value 2.207 > the 5th retrieval result total ranking value 2.052 > the 3rd retrieval result total ranking value 1.901, therefore former sorting position is the second retrieval result of 2, and after rearrangement, rearrangement position is 1;Former sorting position is the first retrieval result of 1, and after rearrangement, rearrangement position is 2; Former sorting position is the 4th retrieval result of 4, and after rearrangement, rearrangement position is 3; Former sorting position is the 5th retrieval result of 5, and after rearrangement, rearrangement position is 4; Former sorting position is the 3rd retrieval result of 3, and after rearrangement, rearrangement position is 5.
Table three:
| Rearrangement position | Former sorting position | Total ranking value |
| 1 | 2 | 2.617 |
| 2 | 1 | 2.401 |
| 3 | 4 | 2.207 |
| 4 | 5 | 2.052 |
| 5 | 3 | 1.901 |
As seen from the above-described embodiment, the retrieval result rearrangement method that the embodiment of the present invention provides, by obtaining retrieval result, according to the temporal information in retrieval result, calculate every time-sequencing value corresponding to retrieval result; The access times of statistics retrieval result, calculate every access times ranking value corresponding to retrieval result according to described access times; According to default target dictionary, calculate in retrieval result, the quantity that in described target dictionary, vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science, and then according to the quantity that vocabulary in described target dictionary occurs, calculates every retrieval industry relevancy ranking value corresponding to retrieval result; Then utilize the former sorting position of retrieval result, calculate name placement value; Described in last weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value, sorts according to described total ranking value. Said process has considered the time in retrieval result, access times, retrieval industry degree of association and the impact on retrieval result of the former sorting position, retrieval industry degree of association is high the most at last, the time is new, user inquires about the maximally related retrieval result of content and comes foremost, it is effectively improved retrieval quality, reduce user's query time, there is significantly high recall precision.
Description by above embodiment of the method, those skilled in the art is it can be understood that can add the mode of required general hardware platform by software to the present invention and realize, hardware can certainly be passed through, but in a lot of situation, the former is embodiment more preferably. Based on such understanding, the part that prior art is contributed by technical scheme substantially in other words can embody with the form of software product, this computer software product is stored in a storage medium, including some instructions with so that a computer equipment (can be personal computer, server, or the network equipment etc.) perform all or part of step of method described in each embodiment of the present invention. And aforesaid storage medium includes: the various media that can store program code such as read only memory (ROM), random access memory (RAM), magnetic disc or CDs.
Corresponding with retrieval result rearrangement method embodiment provided by the invention, present invention also offers a kind of retrieval result rearrangement device.
Referring to Fig. 6, for a kind of structural representation retrieving result rearrangement device that the embodiment of the present invention provides, described device includes:
Retrieval result acquisition module 11, is used for obtaining retrieval result;
Time-sequencing value computing module 12, for according to the temporal information in retrieval result, calculating every time-sequencing value corresponding to retrieval result;
Access times statistical module 13, for adding up the access times of retrieval result;
Access times ranking value computing module 14, for according to described access times, calculating every access times ranking value corresponding to retrieval result;
Vocabulary quantity computing module 15, for according to the target dictionary preset, calculating in retrieval result, the quantity that in described target dictionary, industry vocabulary occurs;Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science;
Retrieval industry relevancy ranking value computing module 16, for the quantity occurred according to industry vocabulary in described target dictionary, calculates every retrieval industry relevancy ranking value corresponding to retrieval result;
Name placement value computing module 17, for utilizing the former sorting position of retrieval result, calculates name placement value;
Total ranking value computing module 18, described in weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value;
Order module 19, for according to described total ranking value, resetting described retrieval result.
Referring to Fig. 7, for the structural representation of a kind of time-sequencing value computing module that the embodiment of the present invention provides, described time-sequencing value computing module 12 also includes:
Reference time acquisition module 121, is used for obtaining reference time information, and described reference time information includes time and month;
Time difference computing module 122, for calculating the time difference of every temporal information corresponding to retrieval result and described reference time information;
Time-sequencing value obtains module 123, for according to described time difference, calculating and obtain described time-sequencing value.
As shown in Figure 8, for the structural representation of a kind of vocabulary quantity computing module that the embodiment of the present invention provides, described vocabulary quantity computing module 15 also includes:
Industry bilingual lexicon acquisition module 151, for obtaining the category of employment information of target dictionary, and according to described category of employment information, selects the industry vocabulary of corresponding category of employment from described target dictionary;
Title vocabulary gauge calculates module 152, retrieves the title in result and described industry vocabulary for comparison, calculates heading remittance number;
Summary vocabulary quantity computing module 153, retrieves the summary in result and described industry vocabulary for comparison, calculates summary vocabulary number.
Referring to Fig. 9, for a kind of structural representation retrieving industry relevancy ranking value computing module that the embodiment of the present invention provides, described retrieval industry relevancy ranking value computing module 16 also includes:
Title coefficient presetting module 161, is used for presetting title coefficient;
Retrieval industry degree of association obtains module 162, is used for utilizing described title coefficient, described heading remittance number and described summary vocabulary number, calculates the retrieval industry degree of association obtaining retrieval result;
Retrieval industry relevancy ranking value obtains module 163, for according to described retrieval industry degree of association, calculating and obtain retrieval industry relevancy ranking value.
For the influence degree that retrieval result is reset by flexible configuration time, access times, retrieval industry degree of association and former sorting position, set up integrated ordered strategy, referring to Figure 10, structural representation for a kind of total ranking value computing module that the embodiment of the present invention provides, the mode that described total ranking value computing module 18 is weighted by calculating obtains described total ranking value, specifically, described total ranking value computing module 18 includes:
Weight coefficient presetting module 181, is used for presetting weight coefficient, and described weight coefficient includes time coefficient, access times coefficient, retrieval industry correlation coefficient and position parameter;
Weighting ranking value computing module 182, for according to the temporal information in described time coefficient and retrieval result, calculating weight temporal ranking value; According to described access times coefficient and access times, calculate weighting access times ranking value; According to the quantity that vocabulary in described retrieval industry correlation coefficient and target dictionary occurs, calculate Weighted Searching industry relevancy ranking value; Former sorting position according to described position parameter and retrieval result, calculates weighting position ranking value;
Accumulator module 183, is used for add up described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and described weighting position ranking value, it is thus achieved that described total ranking value.
As seen from the above-described embodiment, the retrieval result rearrangement device that the embodiment of the present invention provides, by obtaining retrieval result, according to the temporal information in retrieval result, calculate every time-sequencing value corresponding to retrieval result; The access times of statistics retrieval result, calculate every access times ranking value corresponding to retrieval result according to described access times; According to default target dictionary, calculate in retrieval result, the quantity that in described target dictionary, vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science, and then according to the quantity that vocabulary in described target dictionary occurs, calculates every retrieval industry relevancy ranking value corresponding to retrieval result; Then utilize the former sorting position of retrieval result, calculate name placement value; Described in last weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value, sorts according to described total ranking value. Said process has considered the time in retrieval result, access times, retrieval industry degree of association and the impact on retrieval result of the former sorting position, retrieval industry degree of association is high the most at last, the time is new, user inquires about the maximally related retrieval result of content and comes foremost, it is effectively improved retrieval quality, reduce user's query time, there is significantly high recall precision.
For convenience of description, it is divided into various unit to be respectively described with function when describing apparatus above. Certainly, the function of each unit can be realized in same or multiple softwares and/or hardware when implementing the present invention.
Each embodiment in this specification all adopts the mode gone forward one by one to describe, between each embodiment identical similar part mutually referring to, what each embodiment stressed is the difference with other embodiments. Especially for device or system embodiment, owing to it is substantially similar to embodiment of the method, so describing fairly simple, relevant part illustrates referring to the part of embodiment of the method. Apparatus and system embodiment described above is merely schematic, the wherein said unit illustrated as separating component can be or may not be physically separate, the parts shown as unit can be or may not be physical location, namely may be located at a place, or can also be distributed on multiple NE. Some or all of module therein can be selected according to the actual needs to realize the purpose of the present embodiment scheme. Those of ordinary skill in the art, when not paying creative work, are namely appreciated that and implement.
The above is only the specific embodiment of the present invention, makes to skilled artisans appreciate that or realize the present invention. The multiple amendment of these embodiments be will be apparent to one skilled in the art, and generic principles defined herein can without departing from the spirit or scope of the present invention, realize in other embodiments. Therefore, the present invention is not intended to be limited to the embodiments shown herein, and is to fit to the widest scope consistent with principles disclosed herein and features of novelty.
Claims (10)
1. a retrieval result rearrangement method, it is characterised in that comprise the following steps:
Obtain retrieval result;
According to the temporal information in retrieval result, calculate every time-sequencing value corresponding to retrieval result;
The access times of statistics retrieval result;
According to described access times, calculate every access times ranking value corresponding to retrieval result;
According to default target dictionary, calculate in retrieval result, the quantity that in described target dictionary, industry vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science;
According to the quantity that industry vocabulary in described target dictionary occurs, calculate every retrieval industry relevancy ranking value corresponding to retrieval result;
Utilize the former sorting position of retrieval result, calculate name placement value;
Described in weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value;
According to described total ranking value, described retrieval result is reset.
2. retrieval result rearrangement method according to claim 1, it is characterised in that described according to the temporal information in retrieval result, calculates every time-sequencing value corresponding to retrieval result, including:
Obtaining reference time information, described reference time information includes time and month;
Calculate the time difference of every temporal information corresponding to retrieval result and described reference time information;
According to described time difference, calculate and obtain described time-sequencing value.
3. retrieval result rearrangement method according to claim 1, it is characterised in that the target dictionary that described basis is preset, calculates in retrieval result, the quantity that in described target dictionary, vocabulary occurs, including:
Obtain category of employment information, and according to described category of employment information, from described target dictionary, select the industry vocabulary of corresponding category of employment;
The title in result and described industry vocabulary are retrieved in comparison, calculate heading remittance number;
The summary in result and described industry vocabulary are retrieved in comparison, calculate summary vocabulary number.
4. retrieval result rearrangement method according to claim 3, it is characterised in that the described quantity occurred according to vocabulary in described target dictionary, calculates every retrieval industry relevancy ranking value corresponding to retrieval result, including:
Preset title coefficient;
Utilize described title coefficient, described heading remittance number and described summary vocabulary number, calculate the retrieval industry degree of association obtaining retrieval result;
According to described retrieval industry degree of association, calculate and obtain retrieval industry relevancy ranking value.
5. retrieval result rearrangement method according to claim 1, it is characterised in that time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value described in described weighted calculation, obtains total ranking value, including:
Preset weight coefficient, and described weight coefficient includes time coefficient, access times coefficient, retrieval industry correlation coefficient and position parameter;
According to the temporal information in described time coefficient and retrieval result, calculate weight temporal ranking value; According to described access times coefficient and access times, calculate weighting access times ranking value; According to the quantity that vocabulary in described retrieval industry correlation coefficient and target dictionary occurs, calculate Weighted Searching industry relevancy ranking value; Former sorting position according to described position parameter and retrieval result, calculates weighting position ranking value;
Cumulative described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and described weighting position ranking value, it is thus achieved that described total ranking value.
6. a retrieval result rearrangement device, it is characterised in that including:
Retrieval result acquisition module, is used for obtaining retrieval result;
Time-sequencing value computing module, for according to the temporal information in retrieval result, calculating every time-sequencing value corresponding to retrieval result;
Access times statistical module, for adding up the access times of retrieval result;
Access times ranking value computing module, for according to described access times, calculating every access times ranking value corresponding to retrieval result;
Vocabulary quantity computing module, for according to the target dictionary preset, calculating in retrieval result, the quantity that in described target dictionary, industry vocabulary occurs; Described target dictionary is include the dictionary of one or more classification industry vocabulary in electric power, space flight, the energy and medical science;
Retrieval industry relevancy ranking value computing module, for the quantity occurred according to industry vocabulary in described target dictionary, calculates every retrieval industry relevancy ranking value corresponding to retrieval result;
Name placement value computing module, for utilizing the former sorting position of retrieval result, calculates name placement value;
Total ranking value computing module, described in weighted calculation, time-sequencing value, described access times ranking value, described retrieval industry relevancy ranking value and described name placement value, obtain total ranking value;
Order module, for according to described total ranking value, resetting described retrieval result.
7. retrieval result rearrangement device according to claim 6, it is characterised in that described time-sequencing value computing module also includes:
Reference time acquisition module, is used for obtaining reference time information, and described reference time information includes time and month;
Time difference computing module, for calculating the time difference of every temporal information corresponding to retrieval result and described reference time information;
Time-sequencing value obtains module, for according to described time difference, calculating and obtain described time-sequencing value.
8. retrieval result rearrangement device according to claim 6, it is characterised in that described vocabulary quantity computing module includes:
Industry bilingual lexicon acquisition module, is used for obtaining category of employment information, and according to described category of employment information, selects the industry vocabulary of corresponding category of employment from described target dictionary;
Title vocabulary gauge calculates module, retrieves the title in result and described industry vocabulary for comparison, calculates heading remittance number;
Summary vocabulary quantity computing module, retrieves the summary in result and described industry vocabulary for comparison, calculates summary vocabulary number.
9. retrieval result rearrangement device according to claim 8, it is characterised in that described retrieval industry relevancy ranking value computing module includes:
Title coefficient presetting module, is used for presetting title coefficient;
Retrieval industry degree of association obtains module, is used for utilizing described title coefficient, described heading remittance number and described summary vocabulary number, calculates the retrieval industry degree of association obtaining retrieval result;
Retrieval industry relevancy ranking value obtains module, for according to described retrieval industry degree of association, calculating and obtain retrieval industry relevancy ranking value.
10. retrieval result rearrangement device according to claim 6, it is characterised in that described total ranking value computing module includes:
Weight coefficient presetting module, is used for presetting weight coefficient, and described weight coefficient includes time coefficient, access times coefficient, retrieval industry correlation coefficient and position parameter;
Weighting ranking value computing module, for according to the temporal information in described time coefficient and retrieval result, calculating weight temporal ranking value; According to described access times coefficient and access times, calculate weighting access times ranking value; According to the quantity that vocabulary in described retrieval industry correlation coefficient and target dictionary occurs, calculate Weighted Searching industry relevancy ranking value; Former sorting position according to described position parameter and retrieval result, calculates weighting position ranking value;
Accumulator module, is used for add up described weight temporal ranking value, described weighting access times ranking value, described Weighted Searching industry relevancy ranking value and described weighting position ranking value, it is thus achieved that described total ranking value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511008470.3A CN105653661A (en) | 2015-12-29 | 2015-12-29 | Search result re-ranking method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511008470.3A CN105653661A (en) | 2015-12-29 | 2015-12-29 | Search result re-ranking method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105653661A true CN105653661A (en) | 2016-06-08 |
Family
ID=56477126
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511008470.3A Pending CN105653661A (en) | 2015-12-29 | 2015-12-29 | Search result re-ranking method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105653661A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107239497A (en) * | 2017-05-02 | 2017-10-10 | 广东万丈金数信息技术股份有限公司 | Hot content searching method and system |
| CN110688405A (en) * | 2019-08-23 | 2020-01-14 | 上海科技发展有限公司 | Expert recommendation method, device, terminal and medium based on artificial intelligence |
| WO2020019563A1 (en) * | 2018-07-27 | 2020-01-30 | 天津字节跳动科技有限公司 | Search sequencing method and apparatus, electronic device, and storage medium |
| CN111223533A (en) * | 2019-12-24 | 2020-06-02 | 深圳市联影医疗数据服务有限公司 | Medical data retrieval method and system |
| CN111352937A (en) * | 2020-02-14 | 2020-06-30 | 山东省科学院海洋仪器仪表研究所 | Parallel data retrieval method for marine ecological environment monitoring |
| CN111522905A (en) * | 2020-04-15 | 2020-08-11 | 武汉灯塔之光科技有限公司 | Document searching method and device based on database |
| CN113779433A (en) * | 2021-08-16 | 2021-12-10 | 深圳市世强元件网络有限公司 | Search result diversification and equalization searching method and computer equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140588A (en) * | 2007-10-10 | 2008-03-12 | 华为技术有限公司 | Method and apparatus for ordering incidence relation search result |
| CN101630321A (en) * | 2009-08-26 | 2010-01-20 | 中山大学 | On-line article screening method based on data mining (DM) |
| CN102722503A (en) * | 2011-03-31 | 2012-10-10 | 北京百度网讯科技有限公司 | Method and device for sequencing search results |
| CN103186574A (en) * | 2011-12-29 | 2013-07-03 | 北京百度网讯科技有限公司 | Method and device for generating searching result |
| CN103309864A (en) * | 2012-03-07 | 2013-09-18 | 腾讯科技（深圳）有限公司 | Method, device and system for displaying search result |
| CN103324644A (en) * | 2012-03-23 | 2013-09-25 | 日电（中国）有限公司 | Query result diversification method |
| CN104090958A (en) * | 2014-07-04 | 2014-10-08 | 许昌学院 | Semantic information retrieval system and method based on domain ontology |
-
2015
- 2015-12-29 CN CN201511008470.3A patent/CN105653661A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140588A (en) * | 2007-10-10 | 2008-03-12 | 华为技术有限公司 | Method and apparatus for ordering incidence relation search result |
| CN101630321A (en) * | 2009-08-26 | 2010-01-20 | 中山大学 | On-line article screening method based on data mining (DM) |
| CN102722503A (en) * | 2011-03-31 | 2012-10-10 | 北京百度网讯科技有限公司 | Method and device for sequencing search results |
| CN103186574A (en) * | 2011-12-29 | 2013-07-03 | 北京百度网讯科技有限公司 | Method and device for generating searching result |
| CN103309864A (en) * | 2012-03-07 | 2013-09-18 | 腾讯科技（深圳）有限公司 | Method, device and system for displaying search result |
| CN103324644A (en) * | 2012-03-23 | 2013-09-25 | 日电（中国）有限公司 | Query result diversification method |
| CN104090958A (en) * | 2014-07-04 | 2014-10-08 | 许昌学院 | Semantic information retrieval system and method based on domain ontology |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107239497A (en) * | 2017-05-02 | 2017-10-10 | 广东万丈金数信息技术股份有限公司 | Hot content searching method and system |
| WO2020019563A1 (en) * | 2018-07-27 | 2020-01-30 | 天津字节跳动科技有限公司 | Search sequencing method and apparatus, electronic device, and storage medium |
| US11194822B2 (en) | 2018-07-27 | 2021-12-07 | Tianjin Bytedance Technology Co., Ltd. | Search ranking method and apparatus, electronic device and storage medium |
| CN110688405A (en) * | 2019-08-23 | 2020-01-14 | 上海科技发展有限公司 | Expert recommendation method, device, terminal and medium based on artificial intelligence |
| CN111223533A (en) * | 2019-12-24 | 2020-06-02 | 深圳市联影医疗数据服务有限公司 | Medical data retrieval method and system |
| CN111223533B (en) * | 2019-12-24 | 2024-02-13 | 深圳市联影医疗数据服务有限公司 | Medical data retrieval method and system |
| CN111352937A (en) * | 2020-02-14 | 2020-06-30 | 山东省科学院海洋仪器仪表研究所 | Parallel data retrieval method for marine ecological environment monitoring |
| CN111522905A (en) * | 2020-04-15 | 2020-08-11 | 武汉灯塔之光科技有限公司 | Document searching method and device based on database |
| CN113779433A (en) * | 2021-08-16 | 2021-12-10 | 深圳市世强元件网络有限公司 | Search result diversification and equalization searching method and computer equipment |
| WO2023020506A1 (en) * | 2021-08-16 | 2023-02-23 | 深圳市世强元件网络有限公司 | Search method with diversified and equalized search results, and computer device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105653661A (en) | Search result re-ranking method and device | |
| CN100595759C (en) | Method and device for enquire enquiry extending as well as related searching word stock | |
| AU2010236897B2 (en) | System and method for ranking search results within citation intensive document collections | |
| US8832057B2 (en) | Results returned for list-seeking queries | |
| CN105631007A (en) | Industry technical information collecting method and system | |
| CN103593425B (en) | Preference-based intelligent retrieval method and system | |
| CN103838798B (en) | Page classifications system and page classifications method | |
| US20140214825A1 (en) | Systems and methods for identifying documents based on citation history | |
| CN102841946A (en) | Commodity data retrieval sequencing and commodity recommendation method and system | |
| US20150161134A1 (en) | Managing a search | |
| CN106055621A (en) | Log retrieval method and device | |
| CN104298736B (en) | Data acquisition system connection method, device and Database Systems | |
| EP2443546A1 (en) | Generating ranked search results using linear and nonlinear ranking models | |
| CN105912609A (en) | Data file processing method and device | |
| EP2631815A1 (en) | Method and device for ordering search results, method and device for providing information | |
| CN105843841A (en) | Small file storing method and system | |
| CN107180093A (en) | Information search method and device and ageing inquiry word recognition method and device | |
| CN102156711A (en) | Cloud storage based power full text retrieval method and system | |
| US20150302036A1 (en) | Method, system and computer program for information retrieval using content algebra | |
| CN106997390A (en) | A kind of equipment part or parts commodity transaction information search method | |
| CN108182182A (en) | Document matching process, device and computer readable storage medium in translation database | |
| CN103761341A (en) | Information matching method and device | |
| CN104615723B (en) | The determination method and apparatus of query word weighted value | |
| Huang et al. | Improving the relevancy of document search using the multi-term adjacency keyword-order model | |
| Taghva et al. | Effects of similarity metrics on document clustering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160608 |