WO2005033972A1 - 類似率算出装置並びに類似率算出プログラム - Google Patents
類似率算出装置並びに類似率算出プログラム Download PDFInfo
- Publication number
- WO2005033972A1 WO2005033972A1 PCT/JP2004/004451 JP2004004451W WO2005033972A1 WO 2005033972 A1 WO2005033972 A1 WO 2005033972A1 JP 2004004451 W JP2004004451 W JP 2004004451W WO 2005033972 A1 WO2005033972 A1 WO 2005033972A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- technical
- group
- document group
- similarity
- cluster
- Prior art date
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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3346—Query execution using probabilistic model
Definitions
- the present invention relates to a similarity ratio calculation device and a similarity ratio calculation program for comparing similar technical documents to determine similarity.
- the conventional patent map states that it is possible to compare the technology of the same or similar R & D themes using patent documents and to know the overall trend and distribution. By looking at the patent map, it is possible for management to analyze management trends such as market trends, technology trends, trends in entrants and rivals, and future potential.
- the patent map if it is necessary to make a macro comparison between the technical documents A related to company A and the technical documents B related to company B, they belong to the technical documents A and B. They compared each technical literature on a micro scale, and derived macro-level comparisons between technical literature groups.
- Fig. 19 is a diagram showing a conventional comparison situation in which the technical documents included in the technical document A group and the technical documents included in the technical document B group are individually microscopically compared.
- Non-Patent Document 1 The intellectual property evaluation device and the like described in Japanese Patent Application Laid-Open No.
- 2000-34080 include an invention that is pending or registered ⁇ ) — Evaluate the property value of intellectual property
- the data on the execution profit is input, the execution profit input means, the data on the compound interest present rate for each year are input, and the compound interest present rate input means is provided.
- -Input Compounded interest rate calculation means for multiplying the data on compounded interest rate for each ⁇ year by multiplying by.
- the intellectual property price calculating means for calculating the intellectual property value by adding the compounded annual value of the compensation amount for each year calculated by the present value calculating means for each year, and the intellectual property price calculating means
- Output device that outputs the intellectual property value calculated by the intellectual property evaluation device. Intellectual property evaluation method, etc.
- Japanese Patent Laid-Open Publication No. 2000-76062 discloses a system and the like in which first data having a predetermined update interval and second data having an update interval shorter than the first data are used.
- a system for evaluating an evaluation item that can fluctuate over time comprising: (a) a means for creating a first evaluation model in response to input of first data to be sampled; b) applying the first data of the sample object to a first evaluation model to calculate a first evaluation output; and (c—) the second data and the second data of the sample object.
- the system is updated on a yearly or quarterly basis. ⁇ Financial data from balance sheets and profit and loss statements.
- the model static model that conducts corporate reputation, such as the probability of childbirth, and the likelihood of childbirth.
- the second is a relatively short update interval.
- Dell Dynamic model / Le
- Non-Patent Document 1 Japanese Patent Application Laid-open No. Hei 8-: sa081, Japanese Patent Application Laid-Open No. 2003-1703-79-1. : No. Public Bulletin, Tokuhei Hei, 1: 0 7.42 05, Japanese Unexamined Patent Application Publication No. 8-2788, 982, Japanese Unexamined Patent Application Publication No. 11-. , And Japanese Patent Application Laid-Open Publication No. 2000-1-3: 3 Published in IE: ⁇ ).
- technical documents related to company A group A and: related to company B: shiko technical documents.
- Group B There is a request to compare the contents described in the technical literature with the ones in a mark-like manner, but even if it is, the technical literature A—group Technical text toast.
- the technical literatures belonging to each group were compared with each other, and a comparison between the technical literature groups was derived from the operation result of the flag. An inconvenience has occurred.
- Non-Patent Document 1 It is possible to use the same or similar R & D theme, such as: Calculating the relative valuation of each technology based on the entire ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ - ⁇ ⁇ ⁇ ⁇ ⁇ quantitative:, qualitative: specific ⁇ 3 ⁇ 4; the Tokuchi desk - 'not a target of the evaluation of trust and investment, Ru 3 ⁇ 4 a company patent strategy decisive: Les problem, it is not possible to calculate the index of technology ⁇ 'Raw ⁇ . -.
- the ratio of the number of technical documents to the total number of technical documents is calculated as follows: dr, ⁇ ⁇ . ⁇ -ViC ⁇
- the calculated ratio is averaged and the similarity is calculated.
- Weights Creates a 's' dictionary and calculates the arrogance based on the weights.
- Non-Patent Document 1 Putty: Toe: The price of the explosion support software is about 150,000 ⁇ 5: about ⁇ 100,000. Of: to operation;
- Sensation _ sensation A matches.
- Similarity ratio calculation crucible Equipment, similarity ratio calculation program, and similarity calculation # 1 method Provide ⁇ h- ⁇ : Target.
- the similarity ratio is calculated to be 0 only when the first technical document group and the 2nd technical document group are completely different from each other.
- the similarity ratio can be calculated only when the document group is the same as the second technical document group, and a large amount of time-consuming calculation is not required, and the arbitraryness of the analyst is mixed.
- the value of the similarity ratio calculated according to the following is unlikely to be changed:
- the probability of the similarity between the first technical document group and the second technical document group is similar. It is an object of the present invention to provide a similarity ratio calculating device capable of calculating a slope ratio calculation method.
- the case where the total technical literature to be compared is Wei ⁇ --: ::: Even if the similarity ratio is calculated in a relatively short calculation time, it is possible. It is intended to provide a calculation device, a similarity calculation program, and a similarity calculation method.
- the technical literature group can be compared in a similar manner: a similarity ratio calculation device, a similarity ratio calculation program, and a similarity ratio calculation: It is intended to do so.
- similarity calculation devices, similarity calculation programs, and similarity ratios can be easily obtained by investors and general practitioners who need to identify 'corporate value' through intangible property. It is intended to provide a calculation method.
- the present invention provides:, a comparative example.
- the present invention inputs the first technical document group and the second technical document group to be compared '': Oni literature. : "Technology" Enter technical information such as IPC.
- I Technical information input method and 1st technique: Technical literature and 2nd technical literature-Technical-Literature:
- the technical documents containing the technical information described above are each evaluated and the retrieved technical documents are classified into: a cluster for each technical information; a decomposition unit; and a result obtained by performing the cluster decomposition. . all click static number and the second of: technical literature group ⁇ first,:. 2 (..!
- the present invention ratio: input the first technical literature group and the second technical literature group that are the target;; .ru "technique: Skill group input means: and techniques such as -key, -IPC, etc. information, force in ⁇ !. ⁇ information ⁇ Chikarate stage, the Yu ⁇ ) technique:.. surgeon document group and the second technical contained in the technical document group ⁇ Document, Te: the ⁇ mosquitoes and technical information Search for technical documents including ⁇ Technology searched for: Each of the three techniques is divided into clusters for each technical information 1: RU ⁇ ⁇ ⁇ 5- 5- 5- 5- 5-
- the present invention provides a first technique to be compared :: inputting a document group and a second technical document group: inputting a technical document group: means and.
- Gluster's decomposition means The total number of clusters obtained as a result of the ⁇ -tar decomposition:
- Reference group of: Probability raised to the power of the number of technical documents to be extracted ( ⁇ V 0 ⁇ : ⁇ );; Sum of tt sampled E values for each mixed cluster: Similarity ratio calculation to calculate the similarity ratio by dividing 'ta': the budding stage, and the above-mentioned calculated ⁇ similarity ratio 3 ⁇ 4 ⁇ output to the recording means, display means, or communication means: output: power: step : It is characterized by
- the present invention provides a technique for inputting a first technical document group and a second technical document group to be compared ⁇ :. ⁇ Input technical information such as IPC, etc.
- Cluster decomposition is performed for each of the technical information
- Star :: Decomposition hands: Steps: The total number of clusters and the first tea technique, including both the technical literature in the sentence, and the technical literature in the 2.- technical literature group ⁇ &:
- the present invention provides: a comparative example; a first ⁇ technology as an elephant; a technical document group inputting a document group and a second technical document group.
- Techniques for entering technical information such as IPC: Techniques for entering technical information: Dan, and: Included in the first and second technical literature groups and the second technical literature group.
- the technical literature including the technical information is searched, and 3 ⁇ 4a clustering means for performing cluster decomposition on the searched technical literature for each technical information; and the above-mentioned guitar; the total number of clusters obtained as a result of the decomposition _
- the current cluster including both technical documents in the technical literature group and the technical literature group.
- Decomposition - is included in the mixed cluster: Ru technical literature Hajimu the: ride ': Koto first Technology ⁇ : Calculate the expected value for extracting the technical literature of the static group, and calculate the expected value and the technical literature ⁇ of the first technical literature group included in the mixed cluster. The expected value difference is calculated as an arbitrary constant (however,! ⁇ . Negative exponent: Correction ⁇ Calculates the sum of each mixed cluster. Kind of calculating the similarity rate by means of ig rate koto means and shearing
- the present invention is a technology for inputting a technology / document group and a second technical document group.
- . ⁇ .- Input technical information such as IPC : Technical information output stage, 1st: Included in the technical literature group and the second technical literature group.
- the technical group of the technical literature group, and the technical group of the second technical group are examples of the technical literature group of the second technical group.
- the first technical document group and the second technical document group which are composed of technical documents such as patent documents and technical reports, are similar to those for calculating an index for judging technical characteristics.
- the rate calculation device A; 1.
- Technical information input means Technology: Clusters that divide documents into clusters for each piece of technical information ⁇ ?
- a similarity ratio calculating means for calculating, as a similarity ratio, a ratio of the number of mixed clusters including both technical documents in the technical document group of the technical document group, It has a recording means, a table; a means for outputting, and an output means for outputting to the communication means, and so on.Based on the ratio of the total number of clusters and the number of mixed clusters, It is possible to easily calculate an index that indicates the similarity of a given technical content.
- the first correction value that is included in each mixed cluster is included in each mixed cluster before calculating the similarity ratio and is included in each mixed _ cluster:
- the second correction value that takes a value according to the degree of mixing with the technical literature of the first technical literature group and the technical literature of the second technical literature group; and: multiplication,: sum of all mixed rasters , ⁇ : The number of clusters
- the function to calculate the resemblance rate is provided, so the presence of the correction term 1 causes the importance to be included according to the amount of technical literature included in the mixed cluster. It is possible to make a correction that means high, and because of the presence of correction item 2, the closer the ratio of technical documents contained in the mixed cluster to the predetermined amount, the more
- Toseiko 1 and ToTadashi claim 2 a:.
- the similarity ratio calculation means includes: a cluster; a cluster; a bite; a correction value proportional to the number of contributions to the power of ⁇ (however, 0 and ⁇ ); _, The sum is calculated, the function is divided by the total number of clusters to calculate the nesting ratio; therefore, the number of technical documents in the cluster is large.) : , '' Can be calculated.
- the similarity ratio calculation means in the raster technique in the raster is calculated by raising the ⁇ -th power (where 0 and ⁇ ) of the number to the total number of clusters, etc. Since the function to calculate the similarity ratio is provided, it is possible to maintain . ⁇ static ratio 1 -.-. Further, normalization factor and ⁇ es in the surgery number of documents;... Having placed flat 'average value, the technical literature number of Hitoshi ⁇ in all clusters to calculate some amount of Gojun technical literature This is possible
- the similarity ratio calculation means [this, among the-second technical literature group: fern. multiplication - (- where shed ⁇ gamma) to each mix click a correction value proportional: calculating a sum for La data - and, provided the function of ⁇ the similarity rate is divided by the total number of clusters. That is, the similarity ratio calculation means (the number of combinations that retrieve ⁇ or ⁇ technical sentences in the ⁇ group, and ⁇ technical sentences in the ⁇ group) ( ⁇ (mixes the A group with the. S: Ta: Naka.The ability to ⁇ figure and calculate f ⁇ '' to Onoko from the number of '10' technical documents to be retrieved) is included in the A and ⁇ groups included in the mixed cluster.
- the similarity ratio can be corrected to a large bias value, a Zen value, and a small value ⁇ case-: to a large correction value.
- .. normalization factor As a child, ⁇ m in the first technical literature group, _ the first technical literature group ⁇ : ⁇ : The maximum probability of extracting n technical literatures Since the ⁇ value raised to the ⁇ power: (however, '.0)'-is arranged, 0 ⁇ class ratio 3 ⁇ 4 .3 ⁇ 4 is guaranteed within the calculation range of the similarity ratio.
- the similarity ratio calculation handbag includes: the 1st ' ⁇ : dedication group included 1-the number of technical documents ⁇ and the number of technical documents included in the second technical document group ⁇ Revision ratio; ' ⁇ N / M and the result of cluster decomposition are obtained as follows: La 'ta; the number of technical documents in Okina 1's technical literature m ⁇ : Correction value proportional to the power (however, 0 ⁇ ⁇ ) is applied to the mixed quota.
- the composition-ratio of the number of technical documents in Group I, Group I and Group I The more similar the mixture ratio of the technical flaws in each gradus is, the more similar: High :. Calculate _ (approach 1 :) :;
- the ratio of the composition ratio of the technical literature quantity of the ⁇ group to the j mixture ratio of the technical literature inspector When it is large, it is possible to reduce the influence of the similarity ratio on the calculation result.
- similarity ratio calculating means ::: 1: 1: Technical literature! Technical documents mixed with ⁇ and 2.
- expected calculates the expected value and the mixed cluster: first included in: the expected value difference between the number of technical literatures technical document group and:.
- ) is calculated for each mixed cluster, the sum is calculated, and the sum is divided by the total number of clusters to calculate the similarity ratio.
- the similarity ratio calculation-means. The second technology-document group and the second technical document group are mixed.
- Technical text of the document group The probability of extracting the documents is multiplied by the number of technical documents in the cluster-decomposed mixed class. 3 ⁇ 4! Calculate the value of the seasonal meal and calculate the difference between the expected value and the number of technical documents in the technical document group of Section -.1, which is included in the mixed cluster, as the expected value difference.
- the difference which is obtained by dividing the difference by the number of technical documents included in the mixed Gusta J, is- ,; Any constant. (However: 1, 1.
- the sum is calculated for each mixed cod; the divided sum is further divided by the total number of clusters to calculate the similarity ratio.
- FIG. 1 is a whole-body composition diagram of the similarity ratio calculation system according to the present invention.
- FIG. 2 is a block diagram of the similarity calculating device according to the present invention.
- FIG. 3 is a diagram showing 'technology: configuration of documents' included in the technical document A group and the technical document B' group.
- ⁇ 4 indicates similarity display processing-1: Flip chart o
- FIG. 5 is a diagram showing an input screen display: example: for calculating a similarity ratio.
- FIG. 6 is a diagram illustrating a display example of a “similarity / ratio display” screen ⁇ informing the user of the calculated similarity ratio.
- FIG. 7 shows the configuration of each cluster after the technical literature group is solved using the similarity ratio calculating apparatus according to the present invention.
- FIG. 8 is a flowchart showing a process of calculating the similarity ratio. :
- Figure 9 is a chart showing the setting conditions used for calculating the similarity ratio :
- Figure 10 shows that there are many technical documents in mixed cluster 1.
- FIG. 1 A first figure.
- FIG. 11 uses correction term 1 (1).
- Example of calculation of similarity ratio for each case is shown in the table.
- -Fig. 12 shows example of calculation of similarity ratio when correction term 2 (1) is used.
- Chart of ⁇ FIG. 13 is a chart of an example of calculating the similarity ratio, in which the correction term 1 (1) and the correction 2:. :: ( ⁇ ' ⁇ :.;
- Fig. 14 shows the case where the correction term 2 (2) is adopted.
- Similarity ratio calculation example Fig. 15 shows the correction term 1 (1) and the correction term 2 ... (L.2). 13 is a chart of a similarity ratio calculation example in the case of adoption.
- FIG. 16 is a chart showing an example of the calculation of the term value difference in the case where conditions 1 to :: 4 are substituted for (Equation 31 ).
- FIG. 18 is a chart of an example of calculating the similarity ratio when the correction term. 1 (1) and the correction term 2 — (: 3) are adopted.
- Fig. 19 is a diagram showing a microscopic comparison of the technical literature included in the technical literature ⁇ group and the technical literature included in the technical literature B group individually: conventional.
- FIG. 1 is a similarity rate calculation cis beam according to the present invention: is an overall configuration diagram of a. .
- the -similarity ratio calculation system comprises: calculating a similarity ratio from a technical reference database 20 via a communication network 10: necessary: technology.
- Apparatus 3 o Record technical reports, including patent reports of patent applications, patent publications of utility patents, utility model reports, etc., via the communication network 10.
- a technical literature database 20 is provided.
- Communication network 10 is the Internet I etc. Communication: Network, Similarity: Calculation.
- FIG. 2 is a block diagram of the similarity ratio calculating device according to the present invention.
- the similarity ratio calculation device. 3- ⁇ The technical information data is sent to the information transmission / reception unit via a public network or a communication network such as a communication network: 36.
- the transmission / reception means 365 can acquire the technical documents necessary for calculating the similarity rate from the technical document denita: base 2: 0 via the communication network 10.
- the similarity calculating device 30 inputs information related to the technical literatures to be compared (comparison conditions) and the conditions for comparison between the literatures from: a user.
- Means 3 7 0 Includes the function of the technical information input means.
- the similarity ratio calculating device 30 reads various information input through the input means 370 and transmits it to the information processing means 380 described later, or based on an instruction from the information processing means 380.
- Display command is output to LED, etc ..
- Input interface 3 7 1 (It may include the function of technical information input means.) And display information such as images and characters. (It may include the function of the output means: and: Based on the command of the information processing means 3.8.0, the display means 3 A display interface for outputting an image signal is provided with a display interface 37 7 3 (including a function of an output means: may be used). It includes input devices such as tablets as well as mice and mice.
- the similarity ratio calculating device 30 has a recording medium mounting section 3.78 in which the recording medium 377 is removably mounted, and a recording medium “3 717 for each: species detailed information: recording medium and reads Intafue varnish 3 ⁇ 7 9.
- the recording Medium: 377 is represented by: semiconductors such as 'mail', MO, magnetic disk, etc .: magnetic recording, optical recording And the like.
- the similarity ratio calculating device 30 includes: a similarity ratio calculating device.3 ⁇ ⁇ the entire information processing means 380 for performing control; and the information 3 ⁇ 43 ⁇ 4 380 R: OM where various constants are recorded, and the information processing operator 380 processes it.
- Memory 3 8 1 is provided. :
- the information processing means 380 (cluster decomposition means or similarity ratio calculating means) is used by the user to input the information to be compared with the technology to be compared:?: Then, the technical literature necessary for calculating the similarity ratio is obtained from the technical literature database 20 and recorded by the recording means 3 8 4 ⁇ Record: Done 1: Rule_even rate : Calculation : 1? Gram and similarity rate calculation program 3 ⁇ 4 Zui was Doconnection Ru o ⁇ * and it ⁇ possible to realize the function of calculating the similarity rate between technical literature on the,:: similar: constant horn out and displays on the display means 3 7 2 results : Function It is possible to undermine.
- the information processing means 380 (the class decomposition means is included in the claims in the r document, the detailed description of the invention, the brief explanation of drawings, the explanation, the summary, etc .: ('Single :. words, idioms, nouns, verbs, auxiliary verbs, adjectives, -adverbs, postpositions, etc.); Then, each of the searched technical documents is classified for each S operation information.
- information processing means 380 is a bibliographic item; food:: items to be included (classification of IPC, etc., application, 'application number, application name,. C: i
- -Information processing means 380 similarity ratio: calculation means:,-:: cluster decomposition
- ⁇ Extraction ⁇ Realize the function of calculating the similarity ratio between the document groups. It.
- All of these processes are performed by the information processing means': 38: ⁇ ; f twisting....
- the purpose of the present invention can be achieved by executing the processes in a shared manner among a plurality of processing devices. It is.
- the similarity ratio calculation device 30 has a similarity ratio calculation device K 3 ⁇ 0: Processing-related : Various constants ⁇ Connect to a communication device on the network: Attribute information, information : y
- RL Uniform Resource Locators
- Gait '-Nii Information, connection information such as DNS (Domain Name System), information on corporate management.
- DNS Domain Name System
- Various information such as technical information: record information: recordable hard disk etc .: ⁇ means: 3: 3 -4, and :;
- Each peripheral circuit including the clock 390, etc. is connected to the bus 399.
- the peripheral circuits of each are controlled based on the 'program' executed by the information processing means 380. It is possible to realize the function: '.
- the technical information input means is capable of inputting technical information such as a key and a C.
- the output means such as an interface can output the similarity ratio calculated by the similarity ratio calculation means to the recording means, the display means, or the communication means. ..
- the database 20 shown in Figure 1 is stored in :::,
- the above-described similarity ratio calculation device 3.A can be cooled by using various computers such as: a computer, a computer, and a coffee station.
- the computer may be repetitively executed with the function of "net.:talk" to distribute the functions: ,
- the similarity ratio calculating apparatus M according to the present invention M and the similarity ratio of the technical literature calculated using the similarity ratio calculation program ⁇ and ⁇ M 1 ′ technical literature group r. (Technical literature A group).
- the first technical document group (technical document ⁇ : group) and the second “technical document.document group” (technical document) group) have some attribute: a collection of technical documents :.
- the more similar the technical contents described in the first technology the document group ('Technical document A group).
- the second technical document group the technical document B group
- the ratio takes a large value and :: definition: e: v.
- the similarity ratio is calculated. Different: even if the conditions are set, the first technical document group (technical document A-group) and the second technical document group (technical document group) The similarity ratio calculated between-(article B group)-and the third 'technical document group (technical document ... group C) and the fourth technical document group (technical document group D): calculated between In order to be able to directly compare the similarity ratio and ', it is possible to perform :: the operation in which the similarity ratio is within a possible range, and then perform an operation such that 0 ⁇ similarity ratio ⁇ 1. However, the possible range of the similarity ratio is limited to this range. . I
- Figure 3 is a view to showing the configuration of the 1 ⁇ . Murrell technical literature in the technical literature group A and technical literature. B group.
- the technical literature ⁇ ⁇ group is composed of M.
- technical literatures of -A1, .A2 ;: A3, ", :: ', and the technical literature: ⁇ 1 ⁇ , ⁇ 2,: ⁇ 3 :,...: ⁇ ⁇ is composed of ⁇ technical documents.
- FIG. 4 shows the similarity ratio display process.
- the similarity ratio calculating device 30 executes S100- “input image-screen readout 'display.'” As follows: Based on the similarity ratio calculation instruction, based on the similarity ratio calculation, the input screen of various conditions related to the similarity ratio calculation is displayed. Display, read out information from recording means 3 8 4-Based on the displayed information-Similarity ratio. :: Display input screen of conditions required for calculation: Display on means 3 ⁇ 72. : '
- Fig. 5 is a diagram showing a display example of the input: screen ffi for calculating the similarity ratio. ⁇ ; As shown in the figure, the input screen shows the T comparison target. The information that specifies the sentence-collection and the extraction of the second technical document group and the information that specifies the technical information such as the key I2PC are displayed. You. It is possible for users to input various items based on the display screen. In the part for inputting the conditions of cluster decomposition, ⁇ The target of patent gazettes, technical reports, etc .. Specification of documents, full text, only claims, etc .: Setting of target part, cluster of IPC, Kinade, etc. It is possible to input various conditions such as decomposition scale.
- a partial force for inputting a correction method is provided for the purpose of calculating the similarity ratio of the mixed cluster ratio.
- the similarity ratio is corrected based on a value corresponding to the amount of technical literature included in each mixed glass: the user can input a correction condition of whether or not to reject the similarity ratio. It is possible.
- the composition ratio ⁇ _ / of the number of technical documents included in the first technical document ⁇ . And the number of technical documents included in the second technical document group:.
- the first-first technical literature group included in the soil.
- the number of technical / literatures m and the number of technical literature in the second technical literature group n were present in the soil.
- And-The ratio between the composition ratio and the mixture ratio is also taken, but the correction value transferred to ⁇ ⁇ (where ',: ⁇ . ⁇ ) is calculated for each mixed cluster. This is divided by the total number of clusters to correct the similarity ratio, etc., and it is possible to select a correction method according to the “mixture ratio:!” In the technical literature.
- the first technical literature group and the second technical literature group are mixed, and the technology of the first technical literature group is extracted from the technical literature group. Multiplied by the technology 3 ⁇ 4 number of contributions included in the clusters that have been decomposed into clusters: multiplication. Expected value to obtain the technical literature of the first technical literature group is calculated. Mixed. Calculate the difference between the number of technical documents in the first technical document group included in the cluster as the expected value difference: as, and calculate the expected value difference as an optional definite teaching (-low :. Calculate the sum of the negative values of the negative exponents,., And-for each mixed cluster:-,-: multiply this by the number of all clusters to correct the similarity ratio, etc. Expected value difference ; -to- .: It is possible to select a correction method according to the value.
- the information processing means 3.80 is based on the user.
- Kao Search based on the input technical document type (for example, patent document).
- - Identify the source, input from the user and obtain the technical literature group based on the designation of the technical literature group '(for example, A: Company technology ⁇ : Contribution-Group A and B company technical document B group).
- the similarity ratio is calculated in the “similarity ratio calculation process.”
- the calculation devices 3 and 0 are the technical documents acquired from the database 20 (for example, the technical documents A and B: From the technical literature B group), the technical literature containing the common PC and keywords specified by the user is selected and the processing is performed to decompose it for each cluster. -As a result of the cluster decomposition, the technical literature belongs to group A.
- the technical literature and technical literature belonging to technical group B are mixed: A cluster is defined as a mixed Kusuda. In the present invention, among all clusters, a mixed cluster exists.
- the similarity ratio is calculated based on
- Similarity ratio calculation output device 30 was calculated. Similarity ratio is displayed on display means 3 72: Notify to user: Not notified The similarity ratio is displayed on the display means 3 7 2 at S. 06. The calculated similarity rate is displayed on the transmission means 3 6 5 and other communication devices via the communication network 10. The recording medium may be recorded and output to stage 3 84. The recording medium may be transmitted to and output from the recording medium. In: Ta: Huh? ⁇ Record and output to the recording medium 377 via the source 379. : Also, the calculated similarity ratio is output to a printing unit via a pre-printer for printing (not shown). -FIG. 6 is a diagram showing a display example of a similarity ratio display screen that notifies the user of the similarity ratio calculated by the similarity ratio calculation device 30 '.
- the similarity rate display side On the other hand: The information that the user extracted and the technique to extract and specify the group of documents and the technical boat information such as the keyword IPc 'were separated into clusters. Input-information such as scale and correction method at the time is displayed for confirmation.
- the correction term: 3 is set as: .. ri:.: Rq 0; DJ.
- the similarity ratio display screen displays the similarity ratio: calculation: the result, the similarity ratio such as c,, ⁇ , etc. to compensate for the similarity ratio, and the like: the ratio calculation conditions are continuously changed.
- Yes There is a part that displays the contents of the '. -The user looks at the calculated similarity ratio and--freely: sets the conditions for calculating the similarity ratio; It is possible to change. .
- Interest 'for the person slide - de The operation was if the bar, information processing means 3 8 0 calendar clock, 3-9.0 is have time based on the coefficient, the slide bar operation is complete: Han Avoid. Then, the processing performed by the elaborate information processing means 3:80 branches to S104.Then, the similarity ratio is calculated again, and the calculation result of the similarity ratio is displayed on the similarity ratio display screen: Perform the following processing.
- the cluster decomposition of technical literature in the present invention is a macro-comparison between ': the first technical literature group' ( ⁇ -group) and the second technical literature group (group ⁇ ). "Calculation"-Classification of technical documents using keyword KPC etc. when calculating '-'.
- the technical documents of both the first technical document group and the second technical document group are mixed into one group.
- a cluster contains m technical documents belonging to the first technical document group and n technical documents belonging to the second technical document group.
- FIG. 7 shows the similarity ratio calculation according to the present invention: ⁇ j.
- the cluster of the technical literature group is divided using the funeral home. , '
- Patent Document B For each existing restaurant meal: IPC" G06F17Z30 " Contribution A1, ". And _: Includes elements from” Patent Document Bl. "
- a cluster can be composed of each of its attributes.
- -technology such as patent publications; 3 ⁇ 43 ⁇ 4: e: Speaking of, filing date, I pc, etc. .: - ⁇ : .. ⁇ ⁇
- the cluster decomposition method of the information processing means 380 etc. is: the first _ technical literature group and the technical literature included in the second technical literature group. : Technology that contains the entered technical information: The contribution is evaluated, and the retrieved technical documents are decomposed into clusters for each technical information.
- a mixed class is defined as follows:
- the IPC “G06F17r / 30” cluster, shown in Fig. 7, has the following technologies: ⁇ : Patent A1 belonging to group A and technical literature B: belongs to group.
- the non-mixed cluster is defined as follows.
- a cluster in which technical documents belonging to technical document A and technical documents belonging to technical document B do not coexist is defined as a non-mixed raster: Fig. 8 , which indicates the similarity ratio calculation process.
- the processing performed by the information processing means 380 is shown in FIG. 4 j.
- the processing performed by the information processing means 380 is S-20.-. 0, to: branching; S, 2 .. Perform the processing from ⁇ -0 onwards.
- the information processing of the similarity rate calculating device 30-hand “Step 3:80 is ⁇ S 2 0 0." Confuses the group of technical documents A and the group of technical documents B "-v SI 0 2". Acquisition of technical literature " ⁇ :: First technical literature group acquired from the database (for example, company A), first technical literature group and second technical literature group of Company B) Mixed ;; Processing to make a group of documents is performed.
- Correction term 1 is included in the mixed cluster, and the larger the amount of 3 ⁇ 4 technique_mouth and text, the higher the similarity rate is considered to be an important cluster. This is a correction term for correcting the similarity ratio.
- the information processing means 380 corrects the similarity ratio according to the degree of mixing with the technical document A and the technical document B included in the mixed cluster in S2 06 "Set the calculation formula of correction term 2". If the user has input an instruction to that effect, the user selects a formula for the correction term based on the instruction. 'Then, a process of substituting a predetermined formula into the correction term 2 according to the content of the correction is performed.
- the correction term 2 is based on the technical literature included in the mixed cluster ⁇ ) ratio: a predetermined amount-the closer the ratio is, the higher the similarity ratio is considered to be the important cluster- Perform correction: ⁇ z.
- the information processing means 3 8.0 multiplies each correction term of correction term 1, correction term 2, supplementary Calculate sum Put out.
- the similarity ratio is calculated by dividing by all clusters.
- Figure 9 shows the setting conditions used for calculating the similarity ratio.
- Figure 9 shows the first technical texts to be compared: the consecration and the 2nd technical literature; each: The technical literature of the group was decomposed into four clusters.
- 4 is a chart showing the number of technical documents present in .4.
- Basic type 1 When the correction term is not taken into account: Similarity ratio f Basic type 1 ')-. Calculation-example. ⁇ Below, the similarity ratio of the basic type without correction term-this type. 1) An example of calculation will be shown. In the calculation example of the similarity ratio (basic type 1), the similarity ratio of the technical literature is calculated by the mixed cluster extraction method.
- the degree of similarity between the technical content and the degree of similarity is considered to be proportional to the “quantity of mixed clusters”.
- Equation 1 the similarity between the technical and literature groups.
- Equation 1 The similarity ratio calculation method considering mixed clusters is defined as 'mixed cluster extraction method'.
- Equation 1 shown below is the most basic idea. The following: (expression In (1), the ratio of the total number of rasters obtained as a result of cluster decomposition to the number of mixed clusters including the 1st technical literature group and the 2nd technical literature group. ) ::: is calculated with a similarity ratio. Therefore, the method of calculating the ratio of the number of all clusters and the number of mixed clusters is limited to the following (Equation 1).
- the number of mixed clusters is a numerical value that indicates the number of clusters in the second technical literature group: belongs to: technical literature and technology belonging to the second technical literature group: literature is mixed. is there.
- the total number of clusters is a numerical value indicating the total number of clusters in which the technical literature of the first technical literature or the technical literature of the second technical literature group exists.
- the crucial documents included in the first and second technical literature groups are cluster-decomposed using KWA-IPC, etc., and all the decomposed clusters. By calculating the ratio of the number of mixed clusters as the similarity ratio, it is possible to calculate a value serving as a basic part of the similarity ratio between the technical document groups.
- the number of mixed clusters is divided by the number of all clusters.
- Cluster decomposition is performed using keywords and LPCs included in the second and third technical literature groups, and the similarity ratio is calculated based on the ratio of the total number of the decomposed clusters to the number of mixed clusters.
- an index indicating the degree of technical similarity between the technical literature groups can be easily calculated. It was found that the similarity ratio calculated here was in proportion to the degree of similarity between the technical literature groups that we considered as common sense. Also, in the present invention, the value of the similarity ratio to be calculated is within the range of '. O similarity ratio ⁇ 1.
- a first technical document group with more conditions first:. 2 technical publications group: a: a comparison with the similarity rate, a first technical document group a: 3-technical literature group -It is also possible to directly compare the similarity ratio with the comparison. .
- Basic type 2 Calculation example of similarity ratio when considering the correction term: (Basic type 2)-Below is an example of calculating the similarity ratio (Basic type 2 ⁇ ) when considering the correction term: ⁇ . -An example of calculating the similarity ratio (basic type 2) is as follows: The above-described similarity ratio: (basic type 1) 'is calculated by adding correction terms 1 to 3 to the calculation example.
- Equation 1 The most basic (Equation 1) above can be understood from the fact that: For example, many ⁇ 3 ⁇ 4 technology ⁇ : dedication. ⁇ 5> Cluster ⁇ . Technology: ⁇ : Dedicated number-large-small: ; : Due to the drawback that it is not considered., Mixed cluster ⁇ : many-, ⁇ 3 ⁇ 4 technique Even if documents are included, two or two technical documents are not included; no: if. Even if the same similarity ratio is calculated: well, we, common sense. In some cases, the degree of similarity may be different from that of fc.
- FIG. 10 shows a situation in which a large number of technical documents are included in the mixed cluster 1: FIG.
- cluster cluster 3 Another cluster (for example, cluster cluster 3, cluster 4) is an important cluster because there are few technical documents, so I think that it is important to compare it with the contribution of cluster 1. h is small. In a situation like the example in Fig. 10, in the case of ⁇ raster l: r: vs. cluster. '2 The effects of cluster 3 and cluster 4 should be neglected.
- the correction term 1 shown in (Equation 2) is a correction term for calculating the similarity rate according to the technical sentence included in the “mixed cluster”. This correction term 1 is considered to be more important as the amount of technical literature contained in the mixed cluster increases. This is a correction term for correcting the similarity ratio.
- correction term 1 Conversely, mixed cluster included is technology; 3 ⁇ 4:... ⁇ ; Roh * is, small Les -, degree, not considered to be important cluster ': ⁇ book is Teigu a £ Ru, U: This is a correction term that can be used to correct the similarity ratio with a little respect.
- Correction term 1 is included in each mixed class. ⁇ . Calculate the first correction value that takes the shade according to the halo; other possible formulas are av and the correction term.
- the correction term 2 shown in (Equation 2) is based on the technique included in the mixed cluster (the degree of mixing of the technical literature A and the technical literature B (technical literature: ⁇ ⁇ and: Chinese; percentage of the technical literature B):
- the correction term for calculating the rate is:
- Correction term 2 is included in the mixed cluster :: How much do you think that it is an important cluster and the similarity is high? 3 ⁇ 43 ⁇ 4Heavy weight: V is a ⁇ correction term that performs weighting and corrects the similarity rate
- correction term 2 is included in each mixed cluster m :: Hydraulic technique of the second technical literature group.
- the technical literature of the second technical literature group is mixed with 1. This is a correction term for which a correction value of 2 can be calculated. .
- the similarity ratio is calculated by calculating the sum of the correction term 1, the correction term 2, or the correction term 3 for all the mixed clusters.
- the number of technical documents is not biased, it is regarded as an important cluster when the number of technical documents is not biased, and weight is heavy. In this case, it is considered as an insignificant cluster: light: weight: with: suru: ta: no.
- the correction term 3 is a correction term for calculating a similarity ratio by performing an arbitrary weighting when a particular patent classification or keyword is noticed. 'This term is a term that is set individually by those who compare technical literature groups. Therefore, in this case, the constant "1" is substituted without consideration. .
- correction term 1 (1) the similarity ratio ⁇ mixed data: technology included in the raster-To capture a large value according to the value of The number of literatures is multiplied ( ⁇ , 0 ⁇ : ct) to the numerator. Then, as the calculation range of 'similarity rate', 0 ⁇ similarity rate d is guaranteed.
- correction term.1 (1) the normalization factor is placed in the denominator in the formula: .. '
- the “normalization factor” is placed in the denominator of the correction term 1, which makes it possible to guarantee 0 similarity ratio ⁇ 1. Then, as the normalization factor of the correction term 1 (1), the average value of the number of technical documents in all clusters is set as' It is possible to calculate the amount of literature.
- each mixed cluster in the case of this embodiment; :: is-cluster
- Equation 6 when Condition-2 is substituted in Equation 4, will be described below.
- the amount of technical documents included in the cluster is larger than the amount of technical documents included in the other clusters by the calculation process of Equation 6, the amount of the technical documents is similar to: _; Can be reflected in the calculation result of the rate. It represents almost all of the trends in calculating the cluster Yuka similarity ratio-so we can see that this property of cluster 1 works to determine the similarity ratio.
- condition 3 the sum of the amount of technical documents included in the cluster is the same as that in the case of condition 2, but in the situation where the amount of technical documents included in cluster 1 is remarkably large since there is no dark at the time of calculating similar: Other:.. W this; ⁇ is extent when the influence of the amount of technical literature is of condition 2 Ru place students .. Ji never have desired Mr.; I. one
- the value of 0.459 is based on the fact that the amount of technical literature included in cluster 1 is-slightly less than in other clusters: cluster 3; It will be corrected so that it is almost not involved.
- condition 4 the sum of the quantities of technical documents included in the field and the cluster in condition 3 is the same, but the first technical document group included in cluster.1 And the second group of technical documents are ⁇ end: ⁇ even :; Therefore, it is desirable not to calculate the similarity rate large because the number of technical documents included in the mixed cluster is large.
- the correction term (1) ⁇ Section 4: There is a possibility that a part that does not match will be generated.
- the correction-term 2 described below will be useful.
- Figure 11 shows an example of calculation of the correction term 1 (1) adoption :: ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ (( ⁇ . .
- Equation 9 The calculation formula (Equation 9) for the correction term 2 (1) shown below is a configuration that performs correction according to the mixing probability of technical documents in the mixed cluster.
- the probability of taking out the second technical document from the middle of the m (from the group A) and the second technical document group from the second group (the group B) is the power of the probability (however, 0; ⁇ 7) is arranged in the numerator.
- the similarity ratio calculation range is 0 ⁇
- m pieces from the first technique, the 3 ⁇ 4saki group j group, the second technical literature The maximum-value of the probability of taking out n technical contributions from the group (group B) is the ⁇ -th power (where 0 ⁇ ) is placed in the denominator as a normalizing factor.
- the normalization factor may be any term that can guarantee that 0 ⁇ similarity rate ⁇ 1, and the normalization factor shown in (Equation 10) is limited to ⁇ > ⁇ > ⁇ . The following ensures the conditions for setting the index ⁇ .
- the number of technical documents in ⁇ ⁇ : group and 8 groups was randomly extracted from the technical document groups in ⁇ _ group and ⁇ group. ⁇ degree ”, compared to For example, it is necessary to correct the value of the similarity ratio: in this case, set the index ⁇ to 7;
- the number of technical documents in Groups ⁇ ⁇ and ⁇ ⁇ included in the mixed cluster is not close to the distribution when randomly extracted from the technical documents in Groups A and ⁇ B.
- the exponent should be set to ⁇ . ⁇ _ __0 and ⁇ ⁇ 1.
- the denominator as a normalization factor is (the number of combinations from which technical literature is extracted from the group A: X: individual—, group B: y?) / (Group A 'and :: Group B) From the mixture: the number of combinations that take out m + n technical documents) :: is arranged, so that: X, ..y is the similarity ratio calculation range from each combination of numbers that maximizes the denominator -; Possible to guarantee 0 ⁇ similarity rate ⁇ 1.
- the molecular exponent ⁇ is set to ⁇ > 1 .: the number of technical documents of the groups A and B included in the -mixed- raster: It is close to the distribution at the time of random extraction from the technical literature group. In addition, it is possible to disregard the distribution of the technical literatures of the groups A and X ⁇ B .:
- the finger ⁇ _ of the numerator should be set to ..: '. 0 ⁇ 7 ⁇ 1.
- the mixed probability of technical documents included in 1 is calculated as 0.:40.9. ., Similarly, the percentage of technical documents included in cluster 2 is also calculated as., ⁇ ) .4: 0.9.
- the value of the correction term 2 (1) of the mixed cluster 2 is also calculated as 1 and-.
- the value of the correction term 2 (1) is calculated as .1- as shown in the following equation (1.:3), so no correction is made, and the similarity rate is calculated as 0: Is done.
- Equation 14 describes the calculation result of Calculation Example 10 “2 (when Condition 2 is cut down in Expression ⁇ ):”.
- Equation 14 shows a calculation example of the mixing probabilities that constitute the numerator of the correction term 2 (1).
- the normalization factor of the denominator is the maximum value of the mixing probability of the mixed cluster 1
- the normalization factor is calculated as 0.133 as follows.
- the normalization factor of microcluster 2 is also calculated to be 0.448. Normalization factor (condition 3, cluster
- the similarity ratio is calculated as 0..25 by the following calculation.
- Equation 20 the calculation results of Calculation: Example 10-4 (when Condition 4 is substituted into Equation 10).
- the sum of the amount of technical documents included in the case of the condition 3 is the same as that in the case of the condition 3, but the technical documents included in the clusters 1 and 2 are the same.
- the similarity ratio is calculated as follows: 0 ⁇ ; 0— 0 1.
- the similarity ratio calculated by the above (Equation 24).
- the mixing probability of the technical literature is much smaller than the maximum value of the mixing probability when it is extracted from the technical literature group A and the technical literature group.
- the similarity ratio (when condition 4 was substituted) was corrected from 0.459 to 0.001 (when condition 4 was substituted for expression 10).
- Figure 1-2 shows a diagram of an example of similarity ratio calculation when the correction term 2 (1) is adopted (supplementary: calculation results when the conditions 1-to 4 are substituted for (1)).
- the technical literature is well mixed: clusters with high mixing probabilities (clusters with conditions that show a large value :) .
- Correction term 2 (1) Shows a large value of J:
- a clusterer where technical literature is not well mixed a cluster with a condition that indicates a low value of the mixing probability
- the value of the correction term 2 (1) is calculated to be almost “0”, which is a small value.
- the value of the similarity ratio also shows a small value.
- Figure 13 shows an example of calculating the similarity of the correction terms 1 (1) and ⁇ '( ⁇ ⁇ rub angle ⁇ U »(The correction terms 1 (1) and 2 (1) include the conditions 1 to 4_, Calculation result when substitution is made).
- the correction term 1 :: (1) and the correction term: 2_ C 1) are added to calculate the similarity ratio. If it is reflected in the calculation result of the rate, in the case: valid- '.:
- the similarity ratio is calculated using the correction term 1 (1) and the complement E term 2. In addition to correcting the similarity ratio, it is possible to correct the similarity ratio to a small value when the mixture of technical documents is uneven.
- the: j correction of the correction term is sensitive to the degree of mixing in the technical literature.
- the correction term 2 (2) is a correction term for correcting the similarity rate in the mixed cluster by using the technical document ⁇ 'ii yi;.
- the first technical document group (eight: the group:): -. Beauty second _ of Subebun ⁇ group:): Included techniques in the literature number of configuration ratios ⁇ each click 7. .3 ⁇ 4 to: ⁇ : ⁇ ⁇ : technology ⁇ ⁇ The closer the mixing ratio ⁇ m of the number of offerings, the higher the value of the low-grade drug.
- the numerator is the technique of group ⁇ and group :: the number of technical documents
- the composition ratio or the mixing ratio of technical documents in each cluster is small. And the number of technical documents in groups A and B, or the mixture ratio of technical documents in each cluster, whichever is greater, is arranged. And the mixture ratio of technical documents in each cluster is the same: the higher the succession rate is calculated, the closer to 1). Further, the similarity ratio can be calculated to be smaller as the composition ratio of the number of technical documents in the groups A and B and the mixture ratio of the technical documents in each cluster differ. -.
- composition ratio of technical documents in Group A and Group B is calculated as the ratio of the mixed ratio of technical documents in each gluster, so the similarity ratio calculation range.Guaranteed that 0 ⁇ similarity ratio ⁇ 1 Ready to do ⁇ ? :! ⁇
- the index of the molecule is set to 0 and 1: ⁇ 3 ⁇ 4: The ratio of the composition ratio of the technical literature quantity of the ⁇ : group and the ⁇ : group to the mixture ratio of the technical literature in the cluster When the ratio is large, the similarity ratio can be calculated as follows:
- each mixed cluster cluster _rata: 1- and raster '2).
- _ Is the number of technical documents in the first technical document group: (group A).
- the mixture ratio of the two technical documents in the second technical document group (group B): J is .2: 1.
- Equation 29 shows the calculation results of Calculation Example 26-3 (when Condition 3 is substituted into Equation 26).
- composition ratio of technical documents quantity Star 1 mixed Claro mix ratio data 2 is first technical literature group (Alpha group) and the ⁇ because technical document group _ (beta group) And the similarity rate is corrected to be small.
- FIG. 14 shows a chart of i when calculating the correction term 2 (# ⁇ similarity calculation example 1 &. (2). : ⁇ Mixed class in condition 1 and condition 2 ⁇ 1 "and mixed cluster 2 in condition 3 and mixed cluster 2 in condition 3 are as shown in Fig.9. '.
- Example 4 It can be said that it is in a state of mixing 4 mixed technique; ⁇ technique: ⁇ offering ratio is the first technical literature group and the first technique: ⁇ offering group eat * If it is close to the ratio). In this case, the effect is to increase the value of the similarity ratio by increasing the value when the correction term is included.
- the mixed cluster 1 of Condition 3 and the mixed cluster of Condition 4 are in a state where technical documents are not well mixed (the mixed ratio of technical documents in the mixed cluster is The technical literature group of the 2nd technology-document group-the ratio of the number of included technical literature is significantly different from the ratio of the number of included technical literatures). There are i-m. : '
- the correction term 1) and the supplementary year '3 ⁇ 42 (-2)' are substituted for condition 1 in the formula. Since the similarity ratio is calculated according to the mixture ratio, the value of the similarity ratio in the case of substituting the value of ⁇ 0.25 is similar to the case of substituting condition 1 into (Equation 1). The rate is smaller than the value of 0.5, but the expected value is close: good, good. Gouging: It shows well the similarity of the techniques between the two groups .
- correction term 1 (1) and complement 2 ( 2 ) use.
- ⁇ is substituted into the total calculation ⁇ ;
- the amount of penn # contained in the cluster ⁇ mixed ⁇ ratio 3 ⁇ 4: similar analog Rate 3 ⁇ 4: Since the similarity rate is calculated, the similarity rate is substituted into condition (2) using the correction term 1 and the correction term (2) _ from the similarity rate 0.5 of condition 2 (without 3 ⁇ 4f3 ⁇ 4J fe3 ⁇ 4 ⁇ fiber)).
- the similarity ratio in the case of 009 i is calculated as follows: ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ Term 1 and correction 3 ⁇ 4 2 : (2) ⁇ ⁇ ⁇ Calculate the frequency to find the class 1 with a large number of technical documents: ⁇ do ⁇ ; Becomes possible.
- the first technical literature group included in a cluster (: Group A):, technical literature. Number: Quantity M and the technical literature of the second technical literature group (Group B) .Quantity 'and force: A. Group and It is natural to think that the closer to the expected value when randomly extracted from the group: ( ⁇ (M + N)), the better the mixture. 3 ⁇ 4: (The probability ratio shown in the above (Equation 9), Or the third mixture of the mixture ratio shown in (Equation 25) and -1 ⁇ 2: Definition of the condition.)
- the first technical technique group ::.
- (: Group) and the second technical document group (' ⁇ ' group) are mixed.
- Probability of taking out the technical literature ( ⁇ ⁇ ( ⁇ ⁇ )) :, Mixed class j: Included in: Included: Yes: Number of technical documents (m + n ))
- the expected value difference between the number of technical documents in (Group A) _ and the expected value is calculated as:: (Equation: 3 JL)-Reference), and this difference is small- ⁇ , almost (0 is higher. Similarity is higher. -. Perform the calculation to make the correction.
- Equation 31 shows an example of calculating the expected value difference.
- Fig. 16 shows an example of calculating the term. Waiting difference when substituting conditions 1 to 4 in (Equation 31) above.
- Equation 3 1 Calculation result by the above (Equation 3 1): ⁇ ;. et al. ⁇ .: To,: A certain number of technical documents in group A and the number of technical documents in group B contained in Expectations when randomly extracting from groups A and B: The longer the fit, the more the fit. It is a good idea to make the value difference a negative number and place it in the exponent part.
- the expected value difference which is a negative value, is placed in the exponent part by K. This is because when the technical literature of the star has the expected value, the difference between the expected value is equal to 0, and when the index is 0, the correction term value is calculated to be 1; However, because the expected value depends on the size of the mixed cluster, the expected value depends on the size of the mixed cluster, so it is better to divide the expected value difference by the number of technical documents included in the cluster: . '
- FIG. 17 shows an example of calculating the similarity ratio when .DELTA.4 is substituted for .DELTA.t expression "3" when 0 is set.
- the correction term 1) and the correction term: 2 t3)
- the similarity ratio (Arc ;: La'Nuta: Included in .. 1.
- the amount of the distribution is randomized from the groups A and B.
- a correction is made to calculate the similarity ratio higher:: L, so the correction terms, 3 ⁇ 4, and ⁇ Capture term ?. (3); Is similar to the case where the condition l— is substituted by using the ratio.
- condition 2 In the case of condition 2,-, mixed cluster: 1 Y, cluster. 1 2:-to .4 :: compared to and mixed; the number of technical documents included in the current cluster Large: small difference in holding value, mixed cluster Technical literature included in 1 Composition ( ⁇ Effect should be emphasized. If condition 2 is substituted into the formula using correction term 1 (1) and correction term 2 (3), it will be included in the cluster. Calculate the similarity rate based on the number of technical documents and the expected value difference (articles included in a certain cluster 3 ⁇ 4Contribution i (Group A) ⁇ 3 ⁇ 4 Number of documents and second technical documents (Group B ), The amount of similarity is calculated as the closer to the expected value when randomly extracted from Groups A and B, the correction is made.) -(-.3-).
- condition 4 the sum of the amount of technical documents J contained in the rata is the same as in the case of condition 3, but it is included in the mixed cluster and ⁇ ::: .. 2: fe3 ⁇ 4 technical literature Is not particularly large: mixing degree.
- ⁇ the weight of mixed cluster 1 should not be pulled.
- correction term 1 (1) and correction term 2 (.3.): Substitute the total length ⁇ ! ⁇ : Into the cluster: Calculate the similarity ratio.
- ⁇ Input technical information such as IPC, etc.
- the similarity ratio calculating means includes a first correction value that takes a value corresponding to the amount of technical literature included in each mixed cluster and a first correction value included in each mixed cluster.
- the similarity rate is calculated by using the correction term 1 and the correction term 2:-In particular, if the similarity rate is corrected by focusing on the mixed cluster: cluster with large technical literature volume, In both cases, the degree of similarity can be corrected to a small value if the degree of mixing in the technical literature is not uniform.
- the similarity ratio calculating means is in proportion to the power of the individual cluster technical number to the power of ⁇ (however, 0 and ⁇ );
- a function to calculate the sum and divide by the total number of clusters to calculate the similarity ratio is provided:-so, the number of technical documents in the cluster is large, and there is an important raster ⁇ ? This is possible.
- the similarity ratio is calculated by means of the technique in each cluster: the number of ⁇ -powers (however, 0 and c) is normalized by the factor of all clusters, etc.
- .Q— ⁇ similarity ratio ⁇ 1— can be guaranteed.
- the average value of the number of technical documents _ in all clusters was arranged as a normalization factor, the average value of technical documents in all clusters: Can be calculated.
- the similarity ratio calculating means the first technical literature group., The probability of taking out-, and the second technical literature group-n technical literatures from: y-th power: :( 'However, 0 and ⁇ ) Correction value in proportion to ⁇ mixed clusters-::: sum-: _ calculation Then, a function of calculating the similarity ratio by dividing by the total number of clusters is provided.
- the similarity ratio calculation means (m in the group A, and the number of combinations to retrieve the technical literature in the middle of the group) / (with the group A:: 3.
- the function to perform: is included in the mixed cluster: ⁇ group and: ⁇ group: number of technical documents Depending on the remoteness (act of work), it is possible to correct the similarity to a larger correction value if the bias is large, and to correct the similarity to a large correction value if the bias is small.
- the standardization factor is m, from the first technical literature group, m.
- the similarity ratio calculation means the number of technical documents included in the group of technical documents, and the number of technical documents M included in the second technical document group and the technology included in the second technical document group: Ratio,: N / M, and the result of cluster decomposition-obtained: obtained mixed.
- the composition ratio and the mixed ratio of-and-were also taken:
- the correction value proportional to the power of ⁇ : (: 0 ⁇ ) was also applied to each mixed cluster.
- It has a function to calculate the sum total, divide by the total number of clusters, and calculate the similarity ratio, so that the number of technical documents in Groups A and B and the composition of technical documents in each-cluster can be calculated.
- the ratio index between the composition ratio and the mixture ratio to be:> 1
- the technical document vocabulary tt of the groups A and B can be compared with the mixture ratio of the technical documents in the cluster. If the ratio is small, the effect of the warmth: ta will be affected.
- -Dangerous rate The result will not be greatly reflected in the calculation result.
- the composition ratio of the technical literature quantity of the ⁇ group and the: ⁇ group, and the technology of each cluster The similarity ratio is increased or decreased according to the ratio between the literature and the mixture ratio.
- the similarity ratio calculating means is: a technical document group obtained by mixing the technical document group of the -1st technical technique with the technical technique group of the 'th. : technical statement:.. the cluster decomposed £ mixed class included skill 'in data on the probability of retrieving the Document; expected values to retrieve the technical literature of the first technical document group by multiplying the number of surgical literature _ calculated, Included in the expected value and the mixed cluster.
- the difference from the number of documents is calculated as the expected value difference, and the period characteristic ft difference is _arbitrary constant.
- ⁇ -Ai Calculate the sum of the complementary JE values, which are the negative indices of 1 and 2), and divide them by the total number of clusters. According to the setting of the value of ⁇ , it is possible to make a correction that makes the 5 similar-calculation of the expected value difference: the result—: ;:
- the similarity ratio calculating means includes: the first technical document group: the first technical sentence: a group of technical documents; Technology ⁇ -The above-mentioned cluster decomposition was performed on the probability of taking out the contribution: was: mixed.
- the cluster included: The number of technical documents was multiplied to obtain the technical documents of the first technical document group, and the expected value was calculated. Then, the difference between the expected value and the first technique included in the mixed cluster .: ⁇ ; ⁇ group .: Surgery: The difference from the number of documents is calculated as the expected value difference, and the period difference is mixed.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Description
Claims
Priority Applications (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
BRPI0415148-8A BRPI0415148A (pt) | 2003-09-30 | 2004-03-29 | dispositivo de cálculo de similaridade e programa de cálculo de similaridade |
US10/573,778 US20060294060A1 (en) | 2003-09-30 | 2004-03-29 | Similarity calculation device and similarity calculation program |
AU2004277629A AU2004277629A1 (en) | 2003-09-30 | 2004-03-29 | Similarity calculation device and similarity calculation program |
JP2005514348A JPWO2005033972A1 (ja) | 2003-09-30 | 2004-03-29 | 類似率算出装置並びに類似率算出プログラム |
CA002540661A CA2540661A1 (en) | 2003-09-30 | 2004-03-29 | Similarity calculation device and similarity calculation program |
EP04724188A EP1669889A4 (en) | 2003-09-30 | 2004-03-29 | DEVICE AND PROGRAM FOR CALCULATING SIMILARITY |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2003-341904 | 2003-09-30 | ||
JP2003341904 | 2003-09-30 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2005033972A1 true WO2005033972A1 (ja) | 2005-04-14 |
Family
ID=34419250
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2004/004451 WO2005033972A1 (ja) | 2003-09-30 | 2004-03-29 | 類似率算出装置並びに類似率算出プログラム |
Country Status (10)
Country | Link |
---|---|
US (1) | US20060294060A1 (ja) |
EP (1) | EP1669889A4 (ja) |
JP (1) | JPWO2005033972A1 (ja) |
KR (1) | KR20060079792A (ja) |
CN (1) | CN1856788A (ja) |
AU (1) | AU2004277629A1 (ja) |
BR (1) | BRPI0415148A (ja) |
CA (1) | CA2540661A1 (ja) |
RU (1) | RU2344474C2 (ja) |
WO (1) | WO2005033972A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101055580B (zh) * | 2006-04-13 | 2011-10-05 | Lg电子株式会社 | 用于检索文档的系统、方法及用户接口 |
CN111353301A (zh) * | 2020-02-24 | 2020-06-30 | 成都网安科技发展有限公司 | 辅助定密方法及装置 |
Families Citing this family (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070192161A1 (en) * | 2005-12-28 | 2007-08-16 | International Business Machines Corporation | On-demand customer satisfaction measurement |
US8661029B1 (en) * | 2006-11-02 | 2014-02-25 | Google Inc. | Modifying search result ranking based on implicit user feedback |
KR100834292B1 (ko) * | 2006-11-06 | 2008-05-30 | 엔에이치엔(주) | 문서 처리 방법 및 시스템 |
US20100287177A1 (en) * | 2009-05-06 | 2010-11-11 | Foundationip, Llc | Method, System, and Apparatus for Searching an Electronic Document Collection |
US20100287148A1 (en) * | 2009-05-08 | 2010-11-11 | Cpa Global Patent Research Limited | Method, System, and Apparatus for Targeted Searching of Multi-Sectional Documents within an Electronic Document Collection |
US8498974B1 (en) | 2009-08-31 | 2013-07-30 | Google Inc. | Refining search results |
US8364679B2 (en) * | 2009-09-17 | 2013-01-29 | Cpa Global Patent Research Limited | Method, system, and apparatus for delivering query results from an electronic document collection |
US8972391B1 (en) | 2009-10-02 | 2015-03-03 | Google Inc. | Recent interest based relevance scoring |
US20110082839A1 (en) * | 2009-10-02 | 2011-04-07 | Foundationip, Llc | Generating intellectual property intelligence using a patent search engine |
US20110119250A1 (en) * | 2009-11-16 | 2011-05-19 | Cpa Global Patent Research Limited | Forward Progress Search Platform |
US9110971B2 (en) * | 2010-02-03 | 2015-08-18 | Thomson Reuters Global Resources | Method and system for ranking intellectual property documents using claim analysis |
US9623119B1 (en) | 2010-06-29 | 2017-04-18 | Google Inc. | Accentuating search results |
US9026536B2 (en) * | 2010-10-17 | 2015-05-05 | Canon Kabushiki Kaisha | Systems and methods for cluster comparison |
KR20120046671A (ko) * | 2010-11-02 | 2012-05-10 | (주)광개토연구소 | 특허 평가 모델 생성 방법, 특허 평가 방법, 특허 분쟁 예측 모델 생성 방법, 특허 분쟁 예측 정보 생성 방법, 특허 라이센싱 예측 정보 생성 방법, 특허 리스크 헤징 정보 생성 방법 및 시스템 |
KR101255181B1 (ko) * | 2011-03-23 | 2013-04-16 | 강민수 | 특허 분쟁 예측 모델 생성 방법, 그 방법을 실시하는 시스템 및 그 방법이 기록된 기록 매체 |
JP5742506B2 (ja) * | 2011-06-27 | 2015-07-01 | 日本電気株式会社 | 文書類似度算出装置 |
RU2469389C1 (ru) * | 2011-11-08 | 2012-12-10 | Учреждение Российской академии наук Институт системного программирования РАН | Способ интеграции профилей пользователей онлайновых социальных сетей |
US20130159346A1 (en) * | 2011-12-15 | 2013-06-20 | Kas Kasravi | Combinatorial document matching |
CN103514172A (zh) * | 2012-06-20 | 2014-01-15 | 同程网络科技股份有限公司 | 设置搜索引擎关键词的下词方法 |
KR102017746B1 (ko) | 2012-11-14 | 2019-09-04 | 한국전자통신연구원 | 유사도 산출 방법 및 그 장치 |
KR20140078969A (ko) * | 2012-12-18 | 2014-06-26 | (주)광개토연구소 | 특허 괴물 정보를 포함하는 특허 정보 제공 방법 및 그 특허 정보 시스템 |
RU2573951C2 (ru) * | 2013-12-17 | 2016-01-27 | Сергей Анатольевич Головин | Устройство формирования информационно-методических ресурсов кафедры |
WO2016170561A1 (en) * | 2015-04-24 | 2016-10-27 | Nec Corporation | An information processing system and an information processing method for semantic enrichment of text |
KR101724302B1 (ko) * | 2016-10-04 | 2017-04-10 | 한국과학기술정보연구원 | 특허분쟁예보장치 및 그 방법 |
EP3333728A1 (en) * | 2016-12-12 | 2018-06-13 | Tata Consultancy Services Limited | System and method for analyzing research literature for strategic decision making of an entity |
RU2696295C1 (ru) * | 2018-10-31 | 2019-08-01 | Алексей Викторович Морозов | Способ формирования и структурирования электронной базы данных |
CN110826595A (zh) * | 2019-09-29 | 2020-02-21 | 广东美的白色家电技术创新中心有限公司 | 菜谱比较方法、装置及计算机存储介质 |
KR102221355B1 (ko) * | 2020-07-27 | 2021-03-02 | 한국과학기술정보연구원 | 유사 특허 분류방법 및 유사 특허 분류시스템 |
CN112632954A (zh) * | 2020-12-29 | 2021-04-09 | 中译语通科技股份有限公司 | 获取机构技术相似性的方法及装置 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08278982A (ja) | 1995-04-05 | 1996-10-22 | Fuji Electric Co Ltd | 類似語または類似文章の検索方法 |
JPH08287081A (ja) | 1995-04-19 | 1996-11-01 | Fuji Xerox Co Ltd | 類似度付きデータ検索装置 |
JPH1074205A (ja) | 1996-08-30 | 1998-03-17 | Matsushita Electric Ind Co Ltd | 類似名称検索装置 |
US5787420A (en) | 1995-12-14 | 1998-07-28 | Xerox Corporation | Method of ordering document clusters without requiring knowledge of user interests |
JPH1173415A (ja) | 1997-08-27 | 1999-03-16 | Toshiba Corp | 類似文書検索装置及び類似文書検索方法 |
JP2001331527A (ja) | 2000-05-24 | 2001-11-30 | Hitachi Ltd | 類似文書検索方法 |
JP2001337992A (ja) | 2000-05-29 | 2001-12-07 | Mitsubishi Electric Corp | 類似検索システム及び類似検索方法 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5317507A (en) * | 1990-11-07 | 1994-05-31 | Gallant Stephen I | Method for document retrieval and for word sense disambiguation using neural networks |
US6154725A (en) * | 1993-12-06 | 2000-11-28 | Donner; Irah H. | Intellectual property (IP) computer-implemented audit system optionally over network architecture, and computer program product for same |
KR100426382B1 (ko) * | 2000-08-23 | 2004-04-08 | 학교법인 김포대학 | 엔트로피 정보와 베이지안 에스오엠을 이용한 문서군집기반의 순위조정 방법 |
US20020161626A1 (en) * | 2001-04-27 | 2002-10-31 | Pierre Plante | Web-assistant based e-marketing method and system |
-
2004
- 2004-03-29 BR BRPI0415148-8A patent/BRPI0415148A/pt not_active IP Right Cessation
- 2004-03-29 RU RU2006114689/09A patent/RU2344474C2/ru not_active IP Right Cessation
- 2004-03-29 JP JP2005514348A patent/JPWO2005033972A1/ja not_active Withdrawn
- 2004-03-29 WO PCT/JP2004/004451 patent/WO2005033972A1/ja active Search and Examination
- 2004-03-29 CN CNA2004800272246A patent/CN1856788A/zh active Pending
- 2004-03-29 US US10/573,778 patent/US20060294060A1/en not_active Abandoned
- 2004-03-29 CA CA002540661A patent/CA2540661A1/en not_active Abandoned
- 2004-03-29 AU AU2004277629A patent/AU2004277629A1/en not_active Abandoned
- 2004-03-29 EP EP04724188A patent/EP1669889A4/en not_active Withdrawn
- 2004-03-29 KR KR1020067002688A patent/KR20060079792A/ko not_active Application Discontinuation
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08278982A (ja) | 1995-04-05 | 1996-10-22 | Fuji Electric Co Ltd | 類似語または類似文章の検索方法 |
JPH08287081A (ja) | 1995-04-19 | 1996-11-01 | Fuji Xerox Co Ltd | 類似度付きデータ検索装置 |
US5787420A (en) | 1995-12-14 | 1998-07-28 | Xerox Corporation | Method of ordering document clusters without requiring knowledge of user interests |
JPH1074205A (ja) | 1996-08-30 | 1998-03-17 | Matsushita Electric Ind Co Ltd | 類似名称検索装置 |
JPH1173415A (ja) | 1997-08-27 | 1999-03-16 | Toshiba Corp | 類似文書検索装置及び類似文書検索方法 |
JP2001331527A (ja) | 2000-05-24 | 2001-11-30 | Hitachi Ltd | 類似文書検索方法 |
JP2001337992A (ja) | 2000-05-29 | 2001-12-07 | Mitsubishi Electric Corp | 類似検索システム及び類似検索方法 |
Non-Patent Citations (3)
Title |
---|
ARAI K.: "Tokkyo joho bunseki to patent map", THE JOURNAL OF INFORMATION SCIENCE AND TECHNOLOGY ASSOCIATION, vol. 53, no. 1, 1 January 2003 (2003-01-01), pages 16 - 21, XP002986874 * |
SALTON ET AL.: "Information Processing and Management", vol. 24, 1988, ELSEVIER, article "Term-Weighting Approaches in Automatic Text Retrieval", pages: 513 - 523 |
See also references of EP1669889A4 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101055580B (zh) * | 2006-04-13 | 2011-10-05 | Lg电子株式会社 | 用于检索文档的系统、方法及用户接口 |
CN111353301A (zh) * | 2020-02-24 | 2020-06-30 | 成都网安科技发展有限公司 | 辅助定密方法及装置 |
Also Published As
Publication number | Publication date |
---|---|
KR20060079792A (ko) | 2006-07-06 |
RU2006114689A (ru) | 2007-11-20 |
EP1669889A1 (en) | 2006-06-14 |
AU2004277629A1 (en) | 2005-04-14 |
JPWO2005033972A1 (ja) | 2006-12-14 |
BRPI0415148A (pt) | 2006-11-28 |
CN1856788A (zh) | 2006-11-01 |
US20060294060A1 (en) | 2006-12-28 |
RU2344474C2 (ru) | 2009-01-20 |
EP1669889A4 (en) | 2007-10-31 |
CA2540661A1 (en) | 2005-04-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2005033972A1 (ja) | 類似率算出装置並びに類似率算出プログラム | |
Rudkowsky et al. | More than bags of words: Sentiment analysis with word embeddings | |
Blasi et al. | Grammars are robustly transmitted even during the emergence of creole languages | |
Caselli et al. | The road to language learning is not entirely iconic: Iconicity, neighborhood density, and frequency facilitate acquisition of sign language | |
Aguinis et al. | Best-practice recommendations for defining, identifying, and handling outliers | |
De Queiroz | Ernst Mayr and the modern concept of species | |
Ludwick et al. | An improved methodology for advancing nursing research: Factorial surveys | |
Walker et al. | Cross-sensory correspondences: Heaviness is dark and low-pitched | |
Warriner et al. | Affective biases in English are bi-dimensional | |
Lind et al. | Norwegian Words: A lexical database for clinicians and researchers | |
Marelli | Word-Embeddings Italian Semantic Spaces: A semantic model for psycholinguistic research | |
Omran et al. | Transfer learning and sentiment analysis of Bahraini dialects sequential text data using multilingual deep learning approach | |
Iacobucci et al. | How to calculate, use, and report variance explained effect size indices and not die trying | |
Hester et al. | Perceived femininity and masculinity contribute independently to facial impressions. | |
Lawson et al. | Free classification of large sets of everyday objects is more thematic than taxonomic | |
Kirkegaard et al. | Biogeographic ancestry, cognitive ability and socioeconomic outcomes | |
Günther et al. | Trying to make it work: Compositional effects in the processing of compound “nonwords” | |
Wang et al. | When do words matter? Understanding the impact of lexical choice on audience perception using individual treatment effect estimation | |
Kourou et al. | Cohort harmonization and integrative analysis from a biomedical engineering perspective | |
Lancaster et al. | Selective visual attention skills differentially predict decoding and reading comprehension performance across reading ability profiles | |
Albouy et al. | Spectro-temporal acoustical markers differentiate speech from song across cultures | |
Hameau et al. | Effects of semantic neighbourhood density on spoken word production | |
Hassanzadeh et al. | Quantifying semantic similarity of clinical evidence in the biomedical literature to facilitate related evidence synthesis | |
Shamir | UDAT: Compound quantitative analysis of text using machine learning | |
CN109977231B (zh) | 一种基于情感衰变因子的抑郁情绪分析方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200480027224.6 Country of ref document: CN |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): BW GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 1020067002688 Country of ref document: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2006294060 Country of ref document: US Ref document number: 2005514348 Country of ref document: JP Ref document number: 10573778 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2540661 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2004724188 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2292/DELNP/2006 Country of ref document: IN |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2004277629 Country of ref document: AU |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2006114689 Country of ref document: RU |
|
DPEN | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed from 20040101) | ||
ENP | Entry into the national phase |
Ref document number: 2004277629 Country of ref document: AU Date of ref document: 20040329 Kind code of ref document: A |
|
WWP | Wipo information: published in national office |
Ref document number: 2004724188 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: PI0415148 Country of ref document: BR |
|
WWP | Wipo information: published in national office |
Ref document number: 10573778 Country of ref document: US |