CN111428482B - Information identification method and device - Google Patents
Information identification method and device Download PDFInfo
- Publication number
- CN111428482B CN111428482B CN202010224121.XA CN202010224121A CN111428482B CN 111428482 B CN111428482 B CN 111428482B CN 202010224121 A CN202010224121 A CN 202010224121A CN 111428482 B CN111428482 B CN 111428482B
- Authority
- CN
- China
- Prior art keywords
- information
- common sense
- piece
- text
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000012216 screening Methods 0.000 claims abstract description 18
- 238000004590 computer program Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000015654 memory Effects 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 8
- 230000002776 aggregation Effects 0.000 description 6
- 238000004220 aggregation Methods 0.000 description 6
- 239000012634 fragment Substances 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 239000013598 vector Substances 0.000 description 4
- 238000004891 communication Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000006116 polymerization reaction Methods 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
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
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
Abstract
The embodiment of the application discloses an information identification method and device. The method comprises the following steps: acquiring common sense information of at least two persons having the same name information; determining at least two characteristic information in each common sense information; calculating the correlation between each characteristic information in the common sense information and the text description information acquired in advance to obtain local correlation information corresponding to each characteristic information in each common sense information; screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain target content in each common sense information; and determining the person matched with the text description information in the at least two persons according to the target content and the text description information in each piece of common sense information.
Description
Technical Field
The embodiment of the application relates to the field of information processing, in particular to an information identification method and device.
Background
In an application scenario using machine learning identification information, there is also great difficulty in how the current machine distinguishes people with the same name in unstructured text, for example, hundreds of people can be displayed with the same name only by using a search engine; when a person's name appears in a piece of text, it will be difficult for the machine to accurately determine what this name information is specifically referring to.
In the related art, a method for searching for homonyms based on machine learning includes: searching to obtain a related document containing the target person name aiming at the input target person name; respectively extracting character relation characteristic information in each related document, counting the character relation characteristic information in each related document, establishing a character relation graph, and calculating relation strength between the target character name and other character names in the character relation graph; establishing character relation feature vectors for each relevant document according to the character names and the relation strength contained in each relevant document; and clustering all related documents according to the character relation feature vector to obtain a character relation clustering result so as to give a search result at least according to the character relation clustering result.
In practical applications, the accuracy of the above-mentioned method for searching for homonymous characters still needs to be improved.
Disclosure of Invention
In order to solve any technical problem, the embodiment of the application provides an information identification method and an information identification device.
In order to achieve the object of the embodiment of the present application, an embodiment of the present application provides an information identification method, including:
acquiring common sense information of at least two persons having the same name information;
determining at least two characteristic information in each common sense information;
calculating the correlation between each characteristic information in the common sense information and the text description information acquired in advance to obtain local correlation information corresponding to each characteristic information in each common sense information;
screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain target content in each common sense information;
and determining the person matched with the text description information in the at least two persons according to the target content and the text description information in each piece of common sense information.
An information identifying apparatus, comprising:
an acquisition module for acquiring common sense information of at least two persons having the same name information;
a first determining module for determining at least two characteristic information from each common sense information;
the calculation module is used for calculating the correlation between each piece of characteristic information in the common sense information and the text description information acquired in advance to obtain local correlation information corresponding to each piece of characteristic information in each piece of common sense information;
the screening module is used for screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain target content in each common sense information;
and the second determining module is used for determining the person matched with the text description information in the at least two persons according to the target content and the text description information in each piece of common sense information.
A storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method described above when run.
An electronic device comprising a memory having stored therein a computer program and a processor arranged to run the computer program to perform the method described above.
One of the above technical solutions has the following advantages or beneficial effects:
the method comprises the steps of obtaining common sense information of at least two persons with the same name information, determining at least two characteristic information in each common sense information, calculating the correlation between each characteristic information in the common sense information and pre-obtained text description information for each common sense information, obtaining local correlation information corresponding to each characteristic information in each common sense information, screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information, obtaining target content in each common sense information, determining persons matched with the text description information in the at least two persons according to the target content and the text description information in each common sense information, and realizing the purpose of providing an explanatory solution from local to whole.
Additional features and advantages of embodiments of the application will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of embodiments of the application. The objectives and other advantages of embodiments of the application may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
The accompanying drawings are included to provide a further understanding of the technical solution of the embodiments of the present application, and are incorporated in and constitute a part of this specification, illustrate and explain the technical solution of the embodiments of the present application, and not to limit the technical solution of the embodiments of the present application.
FIG. 1 is a flowchart of an information identification method according to an embodiment of the present application;
fig. 2 is a schematic diagram of an information identification method according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a local correlation model according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a multi-stage polymerization model provided by an embodiment of the present application;
fig. 5 is a block diagram of an information identifying apparatus according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the embodiments of the present application will be described in detail hereinafter with reference to the accompanying drawings. It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be arbitrarily combined with each other.
In the process of implementing the present application, the inventor finds that the reason for low recognition accuracy in the related art is that the unsupervised clustering method used in the recognition process can only cluster into a specified number of classes, but cannot explain each class specific designation, so that each name specific designation cannot be determined in the process of homonymous recognition, and therefore, the content lacks of interpretability, and the accuracy cannot be ensured.
Based on the analysis, the embodiment of the application integrates common sense of people into the prior knowledge of the machine for text understanding through the common sense reasoning technology, and the common sense reasoning technology is applied to the identification of the same-name people in the text paragraphs, so that the understanding capability of the machine can be effectively improved, and the problem of the same-name people in the text paragraphs is solved.
Fig. 1 is a flowchart of an information identification method according to an embodiment of the present application. As shown in fig. 1, the method includes:
step 101, obtaining common sense information of at least two persons with the same name information;
in an exemplary embodiment, the general character summary and the basic information in the encyclopedia character knowledge are taken as understanding bases, namely the general knowledge information, wherein the acquisition of the general knowledge information can be acquired from the encyclopedia character knowledge searched on the internet.
In an exemplary embodiment, the common sense information may be at least one of household information, age information, personal life information, and relative information of the person.
Taking the name of the person as Zhang three as an example, the person encyclopedia knowledge records a plurality of pieces of common sense information with the name of Zhang three, and the method can comprise the following steps: the place where the household is located is Zhang san where the A province engages in doctor occupation; the house is Zhang III of C city of B province which is engaged in the teacher occupation; an age of 28 years engages in the actor industry and participates in the three of two television shows.
Step 102, determining at least two characteristic information in each common sense information;
in an exemplary embodiment, the number of features in the different common sense information may be the same or different, and the content of the features may also be different.
In an exemplary embodiment, said determining at least two characteristic information in each common sense information comprises:
acquiring content contained in each piece of common sense information;
and taking the text with the content meeting the judgment condition of the preset similar content as a text segment to obtain at least two text segments in each piece of common sense information, and taking each text segment as a characteristic information.
Taking the above-listed example as an example, the home location is Zhang three, where the doctor career is engaged in the province A, two characteristics can be obtained, namely, "the home location is A province" and "the doctor career is engaged in" respectively.
The judgment condition of the similar content may be judged based on context information between contents expressed by text contents. If the context correlation degree between the two text contents is larger than a preset threshold value, the two text contents are indicated to be similar, otherwise, the two text contents are indicated to be dissimilar.
Taking the above listed examples as examples, the "home location is city C in B province" and "engage in actor industry and participate in two television dramas" can be used as one feature information.
Step 103, calculating the correlation between each characteristic information in the common sense information and the text description information acquired in advance for each common sense information to obtain local correlation information corresponding to each characteristic information in each common sense information;
in an exemplary embodiment, the text description information is a query condition based on which the same name information is filtered, and is used for filtering the task of multiple queried same names by using the text description information, and determining characters matched with the text description information.
In one exemplary embodiment, the degree of importance of each feature in the common sense information is obtained by calculating the local degree of correlation between each feature in the common sense information and the text description information, thereby determining the degree of correlation between each feature in the different common sense information and the text description information, wherein the higher the local degree of correlation between the feature and the text description information is, the higher the importance of the feature is; conversely, the lower the local relevance of a feature to the text description, the lower the importance of that feature.
Step 104, screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain the target content in each common sense information;
in an exemplary embodiment, according to the local relevance information, the content in the common sense information is reduced, the characteristic with high relevance to the text description information is reserved, the characteristic with low relevance to the text description information is removed, the purpose of reserving useful information and deleting useless information is achieved, the target content is obtained, and a more accurate data basis is provided for the specific character to be determined later.
In an exemplary embodiment, the corresponding weight information may be determined for the feature information of each common sense information, and the feature information in each common sense information may be filtered according to the weight information.
And 105, determining the character matched with the text description information in the at least two characters according to the target content and the text description information in each piece of common sense information.
In an exemplary embodiment, compared with the conventional sense information and the text description information which are directly compared in the related art, the method and the device for calculating the text description information by utilizing the target content and the text description information in the step integrally compare, since the content irrelevant to the text description information is deleted from the target content and the content relevant to the text description information is reserved, data support is provided for subsequent accurate judgment, and calculation accuracy is improved.
The method provided by the embodiment of the application comprises the steps of obtaining common sense information of at least two persons with the same name information, determining at least two characteristic information in each common sense information, calculating the correlation between each characteristic information in the common sense information and the pre-obtained text description information for each common sense information, obtaining local correlation information corresponding to each characteristic information in each common sense information, screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information, obtaining target content in each common sense information, determining the person matched with the text description information in the at least two persons according to the target content in each common sense information, realizing the local-to-whole reasoning concept, achieving the aim of providing an interpretable solution, screening the target content from the common sense information by determining the correlation degree of the content in the common sense information for identifying the same name person, and utilizing the target content to determine the characteristic information matched with the text description information, thereby improving the identification precision and the identification efficiency of the person in the absence of the character resume information.
The following describes the method provided by the embodiment of the application:
fig. 2 is a schematic diagram of an information identification method according to an embodiment of the present application. As shown in fig. 2, the method includes:
step 201, receiving an externally input paragraph text, wherein the paragraph text may be a sentence or a paragraph;
step 201, querying common sense knowledge including the name by using the name mentioned in the paragraph text;
wherein, the common sense information can be a abstract (summary) or an information box (infobox) in the encyclopedia information of the name;
step 203, calculating the local correlation between the feature information in each piece of common sense information and the received paragraph text by using a local correlation model;
step 204, utilizing a multi-level aggregation model to combine local correlations in the same common sense information to obtain an overall representation;
step 204, obtaining the character with highest relevance to the paragraph text according to the representation of the whole corresponding to each common sense knowledge.
In the above step, a local correlation model is used to calculate the correlation between each part of the common sense of the person and the text of the paragraph, and further locate the common sense segment with important correlation;
and the multi-level aggregation model is used for aggregating all the part of information of the common sense knowledge according to the weight to obtain the matching degree of the characters and the texts.
In an exemplary embodiment, the local correlation information is determined according to at least one of the following information, including:
calculating the context correlation of each feature information and the text description information;
calculating the text relativity between each word in each feature information and each word in the text feature information;
and determining the attention degree information of each piece of characteristic information and the text characteristic information.
The calculation of the local correlation may be performed by a preset model, and the information obtained in the above embodiment may be processed by a calculation processing layer in the model.
Part of the common sense of the person is related to the input text and part of the content is not related, so that independent reasoning needs to be performed on each part of the content. The specific process is as follows:
α i ＝f e (P i ,Input)
where fe is the function of the decision text implication to be pre-trained. Pi is a segment of common sense of the person and input is the input text.
In order for each part of the common sense content to incorporate context-related information before judging importance, xi after vector representation can be input into a BiLSTM through pi, so that a vector representation containing context-related information can be generated. Wherein BiLSTM is an abbreviation for Bi-directional Long Short-Term Memory, and is formed by combining forward LSTM with backward LSTM. Both are used to model context information in natural language processing tasks.
Fig. 3 is a schematic diagram of a local correlation model according to an embodiment of the present application. As shown in fig. 3, the local correlation model includes:
encoding layers respectively carrying out context-related representation on the common sense fragments and the input text;
cross-attribute laminates, producing a relationship correlation Cross-attribute for each word in the segment and each word in the input text;
other layers, the rest of which are representations that add attention to the segment and the input text according to cross-attribute;
join filters, which generate Xi composed of Xi output by each segment of the next layer and alpha generated by a local correlation module i A weighted combination is made to the representation Y for the entire paragraph of common sense.
The correlation degree corresponding to each characteristic information can be more accurately determined by utilizing the information, and the common sense information can be more accurately embodied
In an exemplary embodiment, the target content in each common sense information is obtained by the following method, including:
according to the local correlation degree information corresponding to each piece of characteristic information in each piece of common sense information, determining the weight value of each piece of characteristic information in each piece of common sense information;
and screening the characteristic information in each piece of common sense information according to the weight value of each piece of characteristic information in each piece of common sense information, and selecting text fragments corresponding to at least two characteristics with the maximum weight value as target contents.
Fig. 4 is a schematic diagram of a multistage polymerization model according to an embodiment of the present application. As shown in fig. 4, the multi-level aggregation model shown in fig. 4 obtains the overall representation of each segment on different levels for the input text through the local relevance model, and fuses the representations of the multiple common sense segments according to the scores in the local relevance model to obtain the paragraph representation of the final entire common sense text.
The multi-stage aggregation model performs the steps of:
a. first, for each segment Pi, the Input text is combined into pairs (Pi, input) and alpha is output through the local correlation module i And respectively inputting the text inclusion function fe corresponding to each fragment.
b. The intermediate representation of n sets is obtained for each fe, where n is the number of fragments that the common sense contains.
c. After fe, the overall representation Y of the segment containing the specific information is obtained, wherein different information can be represented at different join layers, and the multi-level aggregation model is that a plurality of aggregation models are built.
d. Finally, the general knowledge overall representation information Y of different persons is input into a feedforward network to obtain the final logic mapping.
The encyclopedic knowledge is used as the common sense, so that the dependence on priori knowledge is reduced; from the local to the whole reasoning ideas, an interpretable solution can be provided, namely, each segment in encyclopedia knowledge evaluates the correlation degree of the identification of the same-name characters, and the model can improve the identification precision and efficiency when the character resume information is missing.
Fig. 5 is a block diagram of an information identifying apparatus according to an embodiment of the present application. As shown in fig. 5, the apparatus shown in fig. 5 includes:
an acquisition module for acquiring common sense information of at least two persons having the same name information;
a first determining module for determining at least two characteristic information from each common sense information;
the calculation module is used for calculating the correlation between each piece of characteristic information in the common sense information and the text description information acquired in advance to obtain local correlation information corresponding to each piece of characteristic information in each piece of common sense information;
the screening module is used for screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain target content in each common sense information;
and the second determining module is used for determining the person matched with the text description information in the at least two persons according to the target content and the text description information in each piece of common sense information.
In one exemplary embodiment, the first determining module includes:
an acquisition unit configured to acquire content included in each common sense information;
the classification unit is used for taking a text with content meeting the judgment condition of the preset similar content as a text segment to obtain at least two text segments in each piece of common sense information, and taking each text segment as a characteristic information.
In one exemplary embodiment, the calculating module is configured to determine the local correlation information according to at least one of the following information, including:
calculating the context correlation of each feature information and the text description information;
calculating the text relativity between each word in each feature information and each word in the text feature information;
and determining the attention degree information of each piece of characteristic information and the text characteristic information.
In an exemplary embodiment, the filtering module is configured to obtain the target content in each piece of common sense information by:
according to the local correlation degree information corresponding to each piece of characteristic information in each piece of common sense information, determining the weight value of each piece of characteristic information in each piece of common sense information;
and screening the characteristic information in each piece of common sense information according to the weight value of each piece of characteristic information in each piece of common sense information, and selecting text fragments corresponding to at least two characteristics with the maximum weight value as target contents.
The device provided by the embodiment of the application determines at least two characteristic information in each common sense information by acquiring the common sense information of at least two persons with the same name information, calculates the correlation between each characteristic information in the common sense information and the pre-acquired text description information for each common sense information, obtains the local correlation information corresponding to each characteristic information in each common sense information, screens the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information, obtains the target content in each common sense information, determines the person matched with the text description information in the at least two persons according to the target content in each common sense information and the text description information, realizes the reasoning idea from local to whole, achieves the aim of providing an interpretable solution, screens the target content from the common sense information by determining the correlation degree of the content in the common sense information for identifying the same name person, and utilizes the target content to determine the characteristic information matched with the text description information, thereby improving the identification precision and efficiency of the person in the absence of the character resume information.
A storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method of any of the preceding claims when run.
An electronic device comprising a memory having stored therein a computer program and a processor arranged to run the computer program to perform the method as claimed in any one of the preceding claims.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
Claims (6)
1. An information identification method, comprising:
acquiring common sense information of at least two persons having the same name information;
determining at least two characteristic information in each common sense information;
calculating the correlation between each characteristic information in the common sense information and the text description information acquired in advance to obtain local correlation information corresponding to each characteristic information in each common sense information;
screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain target content in each common sense information;
determining the person matched with the text description information in the at least two persons according to the target content and the text description information in each piece of common sense information;
the local correlation information is determined according to at least one of the following information, including:
calculating the context correlation of each feature information and the text description information;
calculating the text relativity between each word in each characteristic information and each word in the text description information;
determining attention degree information of each piece of characteristic information and the text description information;
the target content in each piece of common sense information is obtained by the following way, including:
according to the local correlation degree information corresponding to each piece of characteristic information in each piece of common sense information, determining the weight value of each piece of characteristic information in each piece of common sense information;
screening the feature information in each common sense information according to the weight value of each feature information in each common sense information, selecting at least two features with the maximum weight value, and determining the text segment corresponding to the selected features as target content.
2. The method of claim 1, wherein said determining at least two characteristic information in each common sense information comprises:
acquiring content contained in each piece of common sense information;
and taking the text with the content meeting the judgment condition of the preset similar content as a text segment to obtain at least two text segments in each piece of common sense information, and taking each text segment as a characteristic information.
3. An information identifying apparatus, comprising:
an acquisition module for acquiring common sense information of at least two persons having the same name information;
a first determining module for determining at least two characteristic information from each common sense information;
the calculation module is used for calculating the correlation between each piece of characteristic information in the common sense information and the text description information acquired in advance to obtain local correlation information corresponding to each piece of characteristic information in each piece of common sense information;
the screening module is used for screening the characteristic information of each common sense information according to the local correlation information corresponding to each characteristic information in each common sense information to obtain target content in each common sense information;
the second determining module is used for determining the person matched with the text description information in the at least two persons according to the target content and the text description information in each piece of common sense information;
the computing module is used for determining local relevance information according to at least one of the following information, and comprises the following steps:
calculating the context correlation of each feature information and the text description information;
calculating the text relativity between each word in each characteristic information and each word in the text description information;
determining attention degree information of each piece of characteristic information and the text description information;
the screening module is used for obtaining target content in each piece of common sense information through the following modes:
according to the local correlation degree information corresponding to each piece of characteristic information in each piece of common sense information, determining the weight value of each piece of characteristic information in each piece of common sense information;
screening the feature information in each common sense information according to the weight value of each feature information in each common sense information, selecting at least two features with the maximum weight value, and determining the text segment corresponding to the selected features as target content.
4. The apparatus of claim 3, wherein the first determining module comprises:
an acquisition unit configured to acquire content included in each common sense information;
the classification unit is used for taking a text with content meeting the judgment condition of the preset similar content as a text segment to obtain at least two text segments in each piece of common sense information, and taking each text segment as a characteristic information.
5. A storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method of claim 1 or 2 when run.
6. An electronic device comprising a memory and a processor, characterized in that the memory has stored therein a computer program, the processor being arranged to run the computer program to perform the method of claim 1 or 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010224121.XA CN111428482B (en) | 2020-03-26 | 2020-03-26 | Information identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010224121.XA CN111428482B (en) | 2020-03-26 | 2020-03-26 | Information identification method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111428482A CN111428482A (en) | 2020-07-17 |
| CN111428482B true CN111428482B (en) | 2023-11-24 |
Family
ID=71548826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010224121.XA Active CN111428482B (en) | 2020-03-26 | 2020-03-26 | Information identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111428482B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182420A (en) * | 2013-05-27 | 2014-12-03 | 华东师范大学 | Ontology-based Chinese name disambiguation method |
| CN107766322A (en) * | 2017-08-31 | 2018-03-06 | 平安科技（深圳）有限公司 | Entity recognition method, electronic equipment and computer-readable recording medium of the same name |
| CN109815401A (en) * | 2019-01-23 | 2019-05-28 | 四川易诚智讯科技有限公司 | A kind of name disambiguation method applied to Web people search |
| CN110781670A (en) * | 2019-10-28 | 2020-02-11 | 合肥工业大学 | Chinese place name semantic disambiguation method based on encyclopedic knowledge base and word vector |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9037567B2 (en) * | 2009-04-15 | 2015-05-19 | Vcvc Iii Llc | Generating user-customized search results and building a semantics-enhanced search engine |
-
2020
- 2020-03-26 CN CN202010224121.XA patent/CN111428482B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182420A (en) * | 2013-05-27 | 2014-12-03 | 华东师范大学 | Ontology-based Chinese name disambiguation method |
| CN107766322A (en) * | 2017-08-31 | 2018-03-06 | 平安科技（深圳）有限公司 | Entity recognition method, electronic equipment and computer-readable recording medium of the same name |
| CN109815401A (en) * | 2019-01-23 | 2019-05-28 | 四川易诚智讯科技有限公司 | A kind of name disambiguation method applied to Web people search |
| CN110781670A (en) * | 2019-10-28 | 2020-02-11 | 合肥工业大学 | Chinese place name semantic disambiguation method based on encyclopedic knowledge base and word vector |
Non-Patent Citations (1)
| Title |
|---|
| 文玥琪 ; 周安民 ; .跨社交平台的用户识别方法研究.现代计算机.(第08期),第38-43页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111428482A (en) | 2020-07-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108427738B (en) | Rapid image retrieval method based on deep learning | |
| CN111488426B (en) | Query intention determining method, device and processing equipment | |
| US8457950B1 (en) | System and method for coreference resolution | |
| CN110569496B (en) | Entity linking method, device and storage medium | |
| US20230039496A1 (en) | Question-and-answer processing method, electronic device and computer readable medium | |
| US9009029B1 (en) | Semantic hashing in entity resolution | |
| WO2021146388A1 (en) | Systems and methods for providing answers to a query | |
| CN103678315A (en) | Image processing device, image processing method and electronic equipment | |
| CN114328800A (en) | Text processing method and device, electronic equipment and computer readable storage medium | |
| CN111325033B (en) | Entity identification method, entity identification device, electronic equipment and computer readable storage medium | |
| CN111428482B (en) | Information identification method and device | |
| Vadicamo et al. | Re-ranking via local embeddings: A use case with permutation-based indexing and the nSimplex projection | |
| CN112528021B (en) | Model training method, model training device and intelligent equipment | |
| CN112766288B (en) | Image processing model construction method, device, electronic equipment and readable storage medium | |
| CN114969253A (en) | Market subject and policy matching method and device, computing device and medium | |
| CN115774778A (en) | Resume processing method and device, electronic equipment and readable storage medium | |
| CN110222156B (en) | Method and device for discovering entity, electronic equipment and computer readable medium | |
| CN113704422A (en) | Text recommendation method and device, computer equipment and storage medium | |
| Liu et al. | Exploiting web images for fine-grained visual recognition by eliminating noisy samples and utilizing hard ones | |
| CN117252183B (en) | Semantic-based multi-source table automatic matching method, device and storage medium | |
| CN115795023B (en) | Document recommendation method, device, equipment and storage medium | |
| CN115795184B (en) | RPA-based scene get-on point recommendation method and device | |
| CN115310564B (en) | Classification label updating method and system | |
| CN116341566B (en) | Text deduplication method and device, electronic equipment and storage medium | |
| CN113377893B (en) | POI fusion and POI fusion threshold determining method, device, equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |