CN117633226A - Classification method and device, storage medium and electronic equipment - Google Patents
Classification method and device, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN117633226A CN117633226A CN202311630836.5A CN202311630836A CN117633226A CN 117633226 A CN117633226 A CN 117633226A CN 202311630836 A CN202311630836 A CN 202311630836A CN 117633226 A CN117633226 A CN 117633226A
- Authority
- CN
- China
- Prior art keywords
- evaluation
- target object
- classification
- text
- classification information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000011156 evaluation Methods 0.000 claims abstract description 410
- 230000000875 corresponding effect Effects 0.000 claims description 81
- 230000009466 transformation Effects 0.000 claims description 27
- 238000012545 processing Methods 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 16
- 230000002596 correlated effect Effects 0.000 claims description 5
- 238000013145 classification model Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000008859 change Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a classification method, a classification device, a storage medium and electronic equipment. The method comprises the following steps: acquiring evaluation texts of the target object in different time periods, wherein the different time periods are a plurality of continuous time periods; determining first evaluation classification information of the evaluation text based on the evaluation text for the evaluation text in any time period; determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text in the time period and the first evaluation classification information of the evaluation text; forming an evaluation classification sequence of the target object based on second evaluation classification information respectively corresponding to the target object in a plurality of time periods; and carrying out risk classification on the target object based on the evaluation classification sequence, and determining the risk type of the target object. And carrying out risk classification on the target object through a large number of evaluation texts in a plurality of time periods, so that the comprehensiveness and diversity of the evaluation texts are improved, and the accuracy of risk classification is further improved.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a classification method, apparatus, storage medium, and electronic device.
Background
With the increasing urgent demands of the financial industry for capturing financial events and market public opinion, the monitoring and analysis of public opinion data, especially news data, in combination with marketing and wind control is receiving more and more attention. In financial public opinion analysis, based on deep semantic understanding technology, after non-structural data such as news are analyzed, structured information in aspects such as entity, emotion and the like is obtained.
In the process of realizing the invention, the prior art is found to have at least the following technical problems: news public opinion has unity and sporadic impact analysis accuracy.
Disclosure of Invention
The invention provides a classification method, a classification device, a storage medium and electronic equipment, so as to improve the accuracy of classifying target objects.
According to an aspect of the present invention, there is provided a classification method including:
acquiring evaluation texts of a target object in different time periods, wherein the different time periods are a plurality of continuous time periods;
determining first evaluation classification information of evaluation text based on the evaluation text aiming at the evaluation text in any time period;
determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text in the time period and the first evaluation classification information of the evaluation text;
Forming an evaluation classification sequence of the target object based on second evaluation classification information respectively corresponding to the target object in a plurality of time periods;
and performing risk classification on the target object based on the evaluation classification sequence, and determining the risk type of the target object.
Optionally, before determining the first rating classification information of the rating text based on the rating text, the method further comprises: identifying whether the evaluation text comprises a summary field, if so, extracting the summary text in the evaluation text based on the summary field; or, generating the abstract of the evaluation text based on the abstract text generation model to obtain the abstract text corresponding to the evaluation text;
accordingly, the determining, based on the evaluation text, the first evaluation classification information of the evaluation text includes: first rating classification information of the rating text is determined based on the rating text and/or the summary text.
Optionally, the first evaluation classification information includes a positive category and a negative category; alternatively, the first evaluation classification information includes a classification value, the classification value being positively correlated with a forward extent;
The determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text in the time period and the first evaluation classification information of the evaluation text comprises the following steps: weighting processing is carried out on the basis of the classification value corresponding to the evaluation text and the reliability data of each evaluation text, so that second evaluation classification information of the target object in the time period is obtained; or, weighting processing is carried out on the basis of the numerical value corresponding to the classification category of the evaluation text and the reliability data of each evaluation text, so as to obtain second evaluation classification information of the target object in the time period.
Optionally, the forming the evaluation classification sequence of the target object based on the second evaluation classification information corresponding to the target object in a plurality of time periods respectively includes: and acquiring a plurality of time lengths, and ordering corresponding second evaluation classification information of a plurality of time periods included in the time lengths according to time sequence to form an evaluation classification sequence corresponding to the time lengths.
Optionally, the risk classification of the target object based on the evaluation classification sequence, determining a risk type of the target object includes: performing risk classification on the target object based on the evaluation classification sequence corresponding to each time length to obtain the risk type of the target object under each time length; and determining the target risk type of the target object based on the risk types of the target object corresponding to the time lengths respectively.
Optionally, the evaluation classification sequence includes second evaluation classification information corresponding to the time period, where the second evaluation classification information is a classification class;
the risk classification of the target object based on the evaluation classification sequence, determining a risk type of the target object, includes: determining a class transformation frequency based on the evaluation classification sequence, determining a risk type of the target object based on the class transformation frequency; and/or determining a class transformation frequency based on the frequency data of the positive and negative classes in the evaluation classification sequence, respectively, and determining the risk type of the target object based on the frequency data of the positive and negative classes, respectively, and the class transformation frequency.
Optionally, the risk classification of the target object based on the evaluation classification sequence, determining a risk type of the target object includes: and inputting the evaluation classification sequence into a risk discrimination model to obtain the risk type of the target object.
According to another aspect of the present invention, there is provided a sorting apparatus, comprising:
the text acquisition module is used for acquiring evaluation texts of the target object in different time periods, wherein the different time periods are a plurality of continuous time periods;
The first classification information acquisition module is used for determining first evaluation classification information of the evaluation text based on the evaluation text aiming at the evaluation text in any time period;
a second classification information acquisition module, configured to determine second evaluation classification information of the target object in the time period based on reliability data of the evaluation text and first evaluation classification information of the evaluation text in the time period;
the evaluation classification sequence determining module is used for forming an evaluation classification sequence of the target object based on second evaluation classification information which corresponds to the target object in a plurality of time periods respectively;
and the classification module is used for performing risk classification on the target object based on the evaluation classification sequence and determining the risk type of the target object.
According to another aspect of the present invention, there is provided an electronic apparatus including:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the classification method according to any one of the embodiments of the invention.
According to another aspect of the present invention, there is provided a computer readable storage medium storing computer instructions for causing a processor to execute the classification method according to any embodiment of the present invention.
According to the technical scheme provided by the embodiment of the invention, the number of the evaluation texts is large by acquiring the evaluation texts in a plurality of different time periods, the comprehensiveness and diversity of the evaluation texts are improved, and errors caused by the sporasiveness of a single evaluation text can be avoided. The second evaluation classification information of the target object in the time period is determined through the first evaluation classification information of each evaluation text in the time period, the accuracy of the second evaluation classification information is improved, meanwhile, each evaluation text corresponds to different reliability data, the second evaluation classification information is determined based on the reliability data of the evaluation text and the first evaluation classification information, and the influence of the evaluation file with low reliability on the second evaluation classification information can be reduced. And forming an evaluation classification sequence through the second evaluation classification information corresponding to the time periods, determining the risk type of the target object based on the evaluation classification sequence, and carrying out risk classification on the target object by integrating the second evaluation classification information corresponding to the time periods, thereby improving the classification accuracy.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention or to delineate the scope of the invention. Other features of the present invention will become apparent from the description that follows.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a classification method according to a first embodiment of the present invention;
fig. 2 is a schematic structural diagram of a sorting device according to a second embodiment of the present invention;
fig. 3 is a schematic structural diagram of an electronic device implementing an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
Fig. 1 is a flowchart of a classification method according to an embodiment of the present invention, where the method may be performed by a classification device, which may be implemented in hardware and/or software, and the classification device may be configured in an electronic device, such as a mobile phone, a computer, a server, etc., where the classification method is applicable to classifying a risk of mechanical energy of a target object through an evaluation file of the target object. As shown in fig. 1, the method includes:
S110, acquiring evaluation texts of the target object in different time periods, wherein the different time periods are a plurality of continuous time periods.
S120, determining first evaluation classification information of evaluation texts based on the evaluation texts aiming at the evaluation texts in any time period.
S130, determining second evaluation classification information of the target object in the time period based on reliability data of the evaluation text and first evaluation classification information of the evaluation text in the time period.
And S140, forming an evaluation classification sequence of the target object based on second evaluation classification information respectively corresponding to the target object in a plurality of time periods.
And S150, performing risk classification on the target object based on the evaluation classification sequence, and determining the risk type of the target object.
The target object may be an object for which an enterprise, organization, person is waiting for risk classification. In some embodiments, the target object may be an object that has a financial activity, which may include, but is not limited to, a loan, a deposit, a transfer, and the like. In some embodiments, the target object may be a registration object of a financial institution or a business object of a financial institution. A set of target objects is preset and maintained. The set includes a plurality of target objects and associated entity information for each target object. The associated entity information of the target object includes, but is not limited to, the name, abbreviation, code name, etc. of the target object, which is not limited herein. The set of target objects may be updated periodically or, in the event that a new or deleted target object in the financial institution is detected.
The evaluation text of the target object may be text information such as content including news public opinion including entity information associated with the target object. The method for obtaining the evaluation text of the target object may be: and pulling the evaluation text from the preset network address at fixed time, identifying entity information in the evaluation text, matching the entity information in the evaluation text with the associated entity information of the target object, and determining the evaluation text successfully matched with the associated entity information of the target object as the evaluation text of the target object. In some embodiments, taking a target object as an example of a business, sentence segmentation may be performed on the obtained evaluation text to obtain a plurality of evaluation sentences of the evaluation text, each evaluation sentence is respectively input into a business entity recognition model, and a business entity of each sentence is obtained through recognition. Among them, business entity recognition models include, but are not limited to, BERT models, biLSTM models, CRF models, and the like. And the enterprise entity in the text to be analyzed is identified through the enterprise entity identification model, so that the identification efficiency of the enterprise entity is improved. The entity recognition model can be obtained through training in advance, in the training process, the entity in the input text is marked in advance, and the undetermined parameters in the entity recognition model are adjusted based on the entity obtained through the entity recognition model recognition in the training process and the marked entity until the training completed entity recognition model is obtained.
Optionally, the preset network address includes network addresses of different applications, and correspondingly, the obtained comment text is an evaluation text issued by different applications, so as to improve the comprehensiveness and diversity of the evaluation text and avoid the limitation of the evaluation text issued by a single application. Optionally, the preset network address includes one or more of a country-level information distribution website, a local-level information distribution website, an organization-level information distribution website and a personal information distribution website, and the obtained evaluation text is an evaluation text distributed by different levels of information distribution websites, so as to improve the comprehensiveness and diversity of the evaluation text.
In this embodiment, the evaluation text of the target object included in each of the plurality of time periods is obtained, and the plurality of time periods are continuous time periods, so that a long time period can be formed. By way of example, a time period may be a day, a week, etc. Alternatively, one time period is one day, and a long time period composed of a plurality of time periods may be one week, one month, one year, or the like.
The number of the evaluation texts in any time period is multiple, and the evaluation classification information of the target object is determined through the multiple evaluation texts in each time period. Specifically, the first evaluation classification information of the target object is determined through each evaluation text, further, the second evaluation classification information of the target object in the time period is determined through the first evaluation classification information of the target object corresponding to a plurality of evaluation texts in the time period, so that the singleness and the sporadic property of a single evaluation text can be compensated, and the accuracy can be improved.
Optionally, the evaluation classification information (the first evaluation classification information or the second evaluation classification information) of the target object includes a positive category and a negative category, wherein the positive category characterizes positive and positive evaluation in the evaluation text and the negative category characterizes negative and negative evaluation in the evaluation text.
Optionally, the evaluation classification information (first evaluation classification information or second evaluation classification information) of the target object includes a classification value, which is positively correlated with the degree of forward direction. For example, the rating classification information of the target object may be a classification value of a specific range, for example, the specific range may be 0-1, and the larger the classification value is, the higher the forward rating in the target object in the rating text is, and the smaller the classification value is, the lower the forward rating in the target object in the rating text is.
In some embodiments, a pre-trained evaluation classification model is called, and each evaluation text is input into the evaluation classification model to obtain first evaluation classification information output by the evaluation classification model. By setting the end-to-end evaluation classification model, the evaluation classification process can be simplified, and the evaluation classification efficiency can be improved. The rating classification model may output an identification such as 0 or 1, where 1 characterizes a positive-going category and 0 characterizes a negative-going category. Alternatively, the evaluation classification model may output a classification value.
The evaluation of different sentences on the target object in the same evaluation text can be different, and the evaluation basic tone of the whole evaluation text can be first-come-last-come or first-come-last-go, etc. In order to avoid the problem that evaluation inconsistency of the same evaluation text on the target object causes influence on accuracy of evaluation classification information, abstract text of the evaluation text is extracted, and the abstract text is summarization of the evaluation text and has consistency on evaluation of the target object. Accordingly, determining first rating classification information of the rating text based on the rating text includes: first rating classification information of the rating text is determined based on the rating text and/or the summary text. Specifically, the abstract text is input into the evaluation classification model to obtain first evaluation classification information of the target object, or the evaluation text and the abstract text are input into the evaluation classification model to obtain first evaluation classification information of the target object, and a reference is provided for the evaluation text through the abstract text, so that accuracy of the first evaluation classification information is improved.
On the basis of the above embodiment, before first rating classification information of the rating text is determined based on the rating text, digest text of the rating text is acquired. The method for acquiring the abstract text comprises the following steps: and identifying whether the evaluation text comprises a summary field, and if so, extracting the summary text in the evaluation text based on the summary field. And matching the abstract fields in the evaluation text, if the matching is successful, determining that the abstract fields are included in the evaluation text, and extracting abstract text corresponding to the abstract fields in the evaluation file.
The obtaining mode of the abstract text can also comprise the following steps: and generating the abstract of the evaluation text based on the abstract text generation model to obtain the abstract text corresponding to the evaluation text. And under the condition that the evaluation text does not comprise the abstract field, inputting the evaluation text into the abstract text generation model to obtain the abstract text corresponding to the evaluation text.
And determining second evaluation classification information of the target object in each time period based on the first evaluation classification information respectively corresponding to the multiple evaluation texts of the target object in each time period so as to comprehensively evaluate the target object in the time period.
It will be appreciated that the evaluation texts from different sources have different reliability, and that the reliability of the evaluation information issued by the country-level information issue address is greater than that of the evaluation information issued by the personal-level information issue address, for example. In this embodiment, reliability data of the evaluation text is determined based on the source of the evaluation text, where reliability data of a plurality of sources (for example, preset network addresses) of the evaluation text are preset, and accordingly, the reliability data of the evaluation text is the same as the reliability data of the sources of the evaluation text.
And taking the reliability data of the evaluation text as a weight, and carrying out weighting processing on the first evaluation classification information of the evaluation text in the time period to obtain second evaluation classification information of the target object in the time period.
Taking the first evaluation classification information as a positive class and a negative class as an example, the determining the second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text and the first evaluation classification information of the evaluation text in the time period includes: and weighting based on the numerical value corresponding to the classification category of the evaluation text and the reliability data of each evaluation text to obtain second evaluation classification information of the target object in the time period.
Wherein, the value corresponding to the positive category is greater than the value corresponding to the negative category, and the value corresponding to the positive category may be 1 and the value corresponding to the negative category may be 0.5. Illustratively, the positive category corresponds to a positive number (e.g., may be 1) and the negative category corresponds to a negative number (e.g., may be-1). And carrying out weighting processing on the numerical value corresponding to the classification category through the reliability data of the evaluation text to obtain a weighted numerical value, and determining whether the weighted numerical value is in the numerical value range of the positive category or the numerical value range of the negative category so as to determine the second evaluation classification information. For example, if the weighted value is in the value range of the forward category, the second evaluation classification information of the target object in the time period is the forward category; and if the weighted numerical value is in the numerical range of the negative category, the second evaluation classification information of the target object in the time period is the negative category.
Taking a first evaluation classification information classification value as an example, the determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text and the first evaluation classification information of the evaluation text in the time period includes: and weighting processing is carried out on the basis of the classification value corresponding to the evaluation text and the reliability data of each evaluation text, so as to obtain second evaluation classification information of the target object in the time period. Weighting data obtained by weighting the classification value corresponding to the evaluation text based on the reliability data of the evaluation text is used as second evaluation classification information; or matching the weighted data with the numerical range of the positive category or the numerical range of the negative category, and determining that the second evaluation classification information of the target object in the time period is the positive category or the negative category.
Second evaluation classification information of the target object in each time period is determined based on the above manner. Optionally, second evaluation classification information of the target object in each time period is stored. Correspondingly, the acquiring the evaluation text of the target object in different time periods may be acquiring the evaluation text of the target object in the new time period, determining the second evaluation classification information of the target object in the new time period based on the evaluation text of the target object in the new time period, and acquiring the second evaluation classification information of the target object in a plurality of pre-stored historical time periods. By storing the second evaluation classification information of the target object in the historical time period, repeated processing procedures of the historical evaluation text can be reduced, processing efficiency of the second evaluation classification information of the target object in a plurality of time periods is improved, and risk classification efficiency of the target object is further improved.
And sorting the second evaluation classification information of the target object corresponding to each of the plurality of time periods according to the time sequence to obtain an evaluation classification sequence of the target object. The change trend of the second evaluation classification information of the target object in a long time period formed by a plurality of time periods can be determined through the evaluation classification sequence of the target object so as to classify the risk of the target object.
In some embodiments, the number of time periods is different according to the risk classification requirement, and accordingly, the number of time periods, i.e., the length of time that the time periods are formed, may be determined according to the risk classification requirement. Taking each time period as one day as an example, the time length may be one month, half year, one year, or the like. Optionally, before the evaluation text of the target object is acquired, in response to the triggering operation, an interactive page is displayed, wherein the interactive page can comprise a target object selection control and a time length selection control for the target object to be processed now and the time length. The selected time period may be one or more, and accordingly, an evaluation classification sequence may be determined based on each time period.
Optionally, forming an evaluation classification sequence of the target object based on second evaluation classification information corresponding to the target object in a plurality of time periods respectively includes: and acquiring a plurality of time lengths, and ordering corresponding second evaluation classification information of a plurality of time periods included in the time lengths according to time sequence to form an evaluation classification sequence corresponding to the time lengths.
In some embodiments, a scoring classification information display diagram may also be generated according to the scoring classification sequence, for example, the scoring classification sequence is converted into a graph or a histogram which changes with time, and the graph or the histogram is displayed on an interactive page so as to facilitate the change trend of the scoring classification information for a target object in a plurality of time periods. Correspondingly, in the display page of the grading classification information, different time lengths can be selected to display grading classification information display diagrams corresponding to the selected time lengths so as to meet different display requirements.
And the risk classification is carried out on the target object through the evaluation classification sequence, and the comprehensive classification is carried out by combining the second evaluation classification information of a plurality of time periods, so that compared with the classification mode of a single evaluation text on the target object, the comprehensiveness of the basis information is improved, and errors caused by the singleness and the sporadic property of the evaluation text are avoided.
In some embodiments, risk classification may be performed on the target object based on an evaluation classification sequence corresponding to a time length, so as to obtain a risk type of the target object, that is, a target risk type of the target object. In some embodiments, risk classification may be performed on the target object based on evaluation classification sequences corresponding to a plurality of time lengths, specifically, a risk type corresponding to the evaluation classification sequence corresponding to each time length is determined, and a target risk type of the target object is determined in combination with the risk types corresponding to the plurality of time lengths. Optionally, performing risk classification on the target object based on the evaluation classification sequence, determining a risk type of the target object includes: performing risk classification on the target object based on the evaluation classification sequence corresponding to each time length to obtain the risk type of the target object under each time length; and determining the target risk type of the target object based on the risk types of the target object corresponding to the time lengths respectively. For example, the risk types corresponding to the time lengths may be weighted to obtain the target risk type, where the time length is positively correlated with the corresponding weight, that is, the greater the time length, the greater the weight of the risk type corresponding to the time length. For example, the risk types corresponding to the time lengths may be displayed, and the target risk type of the target object may be determined based on a selection operation of the risk type corresponding to any time length.
On the basis of the above embodiment, performing risk classification on the target object based on the evaluation classification sequence, determining a risk type of the target object includes: and inputting the evaluation classification sequence into a risk discrimination model to obtain the risk type of the target object. The risk discrimination model may be a machine learning model such as a neural network, and the risk discrimination model is trained in advance, for example, may be trained based on an evaluation classification sequence and a risk tag of a sample object.
Optionally, the risk type of the target object includes a risk level type of the target object, for example, the risk level type includes a first risk level, a second risk level, and the like, where the risk levels corresponding to the first risk level and the second risk level gradually decrease.
Optionally, the risk type of the target object includes a risk type and a stability type. Wherein the risk degree corresponding to the risk type is higher than the risk degree corresponding to the stable type.
On the basis of the above embodiment, the determination of the risk type of the target object based on the evaluation classification sequence may be based on the fluctuation state or the change trend determination of the second evaluation classification information in the evaluation classification sequence, and the greater the fluctuation degree of the second evaluation classification information in the evaluation classification sequence, the lower the risk level of the target object and/or the higher the risk level of the target object, the more the risk type of the target object tends to the risk type. Illustratively, the change trend of the second evaluation classification information in the evaluation classification sequence is that the fluctuation of the second evaluation classification information is larger and larger, or the second evaluation classification information is changed from the positive category to the negative category, the lower the risk level of the target object is, and/or the higher the risk level of the target object is, the more the risk type of the target object tends to be the risk type.
Taking the evaluation classification sequence as an example, the evaluation classification sequence includes second evaluation classification information corresponding to a plurality of time periods, the second evaluation classification information is classified into classification categories, the risk classification is performed on the target object based on the evaluation classification sequence, and the risk type of the target object is determined, including: determining a class transformation frequency based on the evaluation classification sequence, determining a risk type of the target object based on the class transformation frequency; wherein, the classification categories corresponding to adjacent time periods are different, which indicates that a category transformation exists, and the transformation frequency of the classification sequence is statistically evaluated to determine the category transformation frequency. The higher the class-change frequency, the greater the degree of fluctuation. The frequency ranges corresponding to different risk types can be preset, category transformation frequencies corresponding to the evaluation classification sequences are matched with the frequency ranges corresponding to different risk types, and the frequency ranges to which the category transformation frequencies corresponding to the evaluation classification sequences belong are determined so as to determine the risk types of the target object. Illustratively, the minimum value of the frequency range corresponding to the risk type is greater than the maximum value of the frequency range corresponding to the stable type.
In some embodiments, when the evaluation classification sequence includes a large number of negative categories, the category transformation frequency is small, but since the large number of negative categories indicate that the overall evaluation of the target object tends to be negative, a certain risk exists, in order to avoid the situation that the risk classification is inaccurate due to single basis of the category transformation frequency, the frequency data of the positive category and the frequency data of the negative category in the evaluation classification sequence can be counted, the risk type of the target object is determined together by the frequency of each category and the category transformation frequency, and the accuracy of risk type determination of the target object is improved.
Optionally, the risk classification of the target object based on the evaluation classification sequence, determining a risk type of the target object includes: and determining a class transformation frequency based on the frequency data of the positive class and the negative class in the evaluation classification sequence, and determining the risk type of the target object based on the frequency data of the positive class and the negative class and the class transformation frequency.
The method comprises the steps of carrying out weighting processing on positive-direction class frequency data, negative-direction class frequency data and class transformation frequency to obtain weighted data, matching the weighted data with frequency ranges corresponding to different risk types, and determining the frequency range to which the class transformation frequency corresponding to the evaluation classification sequence belongs to so as to determine the risk type of the target object. The weight of the frequency data of the positive category may be a positive number, the weight of the frequency data of the negative category may be a negative number, and the weight of the category transformation frequency may be a negative number.
According to the technical scheme, the number of the evaluation texts is large by acquiring the evaluation texts in a plurality of different time periods, the comprehensiveness and diversity of the evaluation texts are improved, and errors caused by sporasiveness of a single evaluation text can be avoided. The second evaluation classification information of the target object in the time period is determined through the first evaluation classification information of each evaluation text in the time period, the accuracy of the second evaluation classification information is improved, meanwhile, each evaluation text corresponds to different reliability data, the second evaluation classification information is determined based on the reliability data of the evaluation text and the first evaluation classification information, and the influence of the evaluation file with low reliability on the second evaluation classification information can be reduced. And forming an evaluation classification sequence through the second evaluation classification information corresponding to the time periods, determining the risk type of the target object based on the evaluation classification sequence, and carrying out risk classification on the target object by integrating the second evaluation classification information corresponding to the time periods, thereby improving the classification accuracy.
Example two
Fig. 2 is a schematic structural diagram of a sorting device according to a second embodiment of the present invention. As shown in fig. 2, the apparatus includes:
a text obtaining module 210, configured to obtain evaluation text of the target object in different time periods, where the different time periods are a plurality of continuous time periods;
a first classification information obtaining module 220, configured to determine, for an evaluation text within any time period, first evaluation classification information of the evaluation text based on the evaluation text;
a second classification information obtaining module 230, configured to determine second evaluation classification information of the target object in the time period based on reliability data of the evaluation text and the first evaluation classification information of the evaluation text in the time period;
an evaluation classification sequence determining module 240, configured to form an evaluation classification sequence of the target object based on second evaluation classification information corresponding to the target object in a plurality of time periods respectively;
and the classification module 250 is used for performing risk classification on the target object based on the evaluation classification sequence and determining the risk type of the target object.
According to the technical scheme, the number of the evaluation texts is large by acquiring the evaluation texts in a plurality of different time periods, the comprehensiveness and diversity of the evaluation texts are improved, and errors caused by sporasiveness of a single evaluation text can be avoided. The second evaluation classification information of the target object in the time period is determined through the first evaluation classification information of each evaluation text in the time period, the accuracy of the second evaluation classification information is improved, meanwhile, each evaluation text corresponds to different reliability data, the second evaluation classification information is determined based on the reliability data of the evaluation text and the first evaluation classification information, and the influence of the evaluation file with low reliability on the second evaluation classification information can be reduced. And forming an evaluation classification sequence through the second evaluation classification information corresponding to the time periods, determining the risk type of the target object based on the evaluation classification sequence, and carrying out risk classification on the target object by integrating the second evaluation classification information corresponding to the time periods, thereby improving the classification accuracy.
On the basis of the above embodiment, optionally, the first classification information obtaining module 220 is further configured to:
before first evaluation classification information of the evaluation text is determined based on the evaluation text, whether a summary field is included in the evaluation text is identified, if yes, the summary text in the evaluation text is extracted based on the summary field; or, generating the abstract of the evaluation text based on the abstract text generation model to obtain the abstract text corresponding to the evaluation text; determining first rating classification information of the rating text based on the rating text and/or the summary text
On the basis of the above embodiment, optionally, the first evaluation classification information includes a positive category and a negative category; alternatively, the first evaluation classification information includes a classification value, the classification value being positively correlated with a forward extent;
the second classification information acquisition module 230 is further configured to: weighting processing is carried out on the basis of the classification value corresponding to the evaluation text and the reliability data of each evaluation text, so that second evaluation classification information of the target object in the time period is obtained; or, weighting processing is carried out on the basis of the numerical value corresponding to the classification category of the evaluation text and the reliability data of each evaluation text, so as to obtain second evaluation classification information of the target object in the time period.
Based on the above embodiment, optionally, the evaluation classification sequence determining module 240 is further configured to: and acquiring a plurality of time lengths, and ordering corresponding second evaluation classification information of a plurality of time periods included in the time lengths according to time sequence to form an evaluation classification sequence corresponding to the time lengths.
Based on the above embodiment, optionally, the classification module 250 is specifically configured to: performing risk classification on the target object based on the evaluation classification sequence corresponding to each time length to obtain the risk type of the target object under each time length; and determining the target risk type of the target object based on the risk types of the target object corresponding to the time lengths respectively.
Optionally, the evaluation classification sequence includes second evaluation classification information corresponding to the time period, where the second evaluation classification information is a classification class;
the classification module 250 is further configured to: determining a class transformation frequency based on the evaluation classification sequence, determining a risk type of the target object based on the class transformation frequency; and/or determining a class transformation frequency based on the frequency data of the positive and negative classes in the evaluation classification sequence, respectively, and determining the risk type of the target object based on the frequency data of the positive and negative classes, respectively, and the class transformation frequency.
Optionally, the classification module 250 is further configured to: and inputting the evaluation classification sequence into a risk discrimination model to obtain the risk type of the target object.
The classification device provided by the embodiment of the invention can execute the classification method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
Example III
Fig. 3 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.
As shown in fig. 3, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.
Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as classification methods.
In some embodiments, the classification method may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the classification method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the classification method in any other suitable way (e.g., by means of firmware).
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
The computer program used to implement the classification method of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
Example IV
The fourth embodiment of the present invention also provides a computer readable storage medium storing computer instructions for causing a processor to execute a classification method, the method comprising:
acquiring evaluation texts of a target object in different time periods, wherein the different time periods are a plurality of continuous time periods; determining first evaluation classification information of evaluation text based on the evaluation text aiming at the evaluation text in any time period; determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text in the time period and the first evaluation classification information of the evaluation text; forming an evaluation classification sequence of the target object based on second evaluation classification information respectively corresponding to the target object in a plurality of time periods; and performing risk classification on the target object based on the evaluation classification sequence, and determining the risk type of the target object.
In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.
The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.
The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method of classification, comprising:
acquiring evaluation texts of a target object in different time periods, wherein the different time periods are a plurality of continuous time periods;
determining first evaluation classification information of evaluation text based on the evaluation text aiming at the evaluation text in any time period;
determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text in the time period and the first evaluation classification information of the evaluation text;
forming an evaluation classification sequence of the target object based on second evaluation classification information respectively corresponding to the target object in a plurality of time periods;
and performing risk classification on the target object based on the evaluation classification sequence, and determining the risk type of the target object.
2. The method of claim 1, wherein prior to determining the first rating classification information for the rating text based on the rating text, the method further comprises:
identifying whether the evaluation text comprises a summary field, if so, extracting the summary text in the evaluation text based on the summary field; or, generating the abstract of the evaluation text based on the abstract text generation model to obtain the abstract text corresponding to the evaluation text;
Accordingly, the determining, based on the evaluation text, the first evaluation classification information of the evaluation text includes:
first rating classification information of the rating text is determined based on the rating text and/or the summary text.
3. The method of claim 1, wherein the first rating classification information comprises a positive category and a negative category; alternatively, the first evaluation classification information includes a classification value, the classification value being positively correlated with a forward extent;
the determining second evaluation classification information of the target object in the time period based on the reliability data of the evaluation text in the time period and the first evaluation classification information of the evaluation text comprises the following steps:
weighting processing is carried out on the basis of the classification value corresponding to the evaluation text and the reliability data of each evaluation text, so that second evaluation classification information of the target object in the time period is obtained; or,
and weighting based on the numerical value corresponding to the classification category of the evaluation text and the reliability data of each evaluation text to obtain second evaluation classification information of the target object in the time period.
4. The method according to claim 1, wherein the forming the evaluation classification sequence of the target object based on the second evaluation classification information respectively corresponding to the target object in a plurality of time periods includes:
And acquiring a plurality of time lengths, and ordering corresponding second evaluation classification information of a plurality of time periods included in the time lengths according to time sequence to form an evaluation classification sequence corresponding to the time lengths.
5. The method of claim 4, wherein risk classifying the target object based on the evaluation classification sequence, determining a risk type of the target object, comprises:
performing risk classification on the target object based on the evaluation classification sequence corresponding to each time length to obtain the risk type of the target object under each time length;
and determining the target risk type of the target object based on the risk types of the target object corresponding to the time lengths respectively.
6. The method of claim 1, wherein the sequence of rating classifications includes corresponding second rating classification information for a time period, the second rating classification information being a classification category;
the risk classification of the target object based on the evaluation classification sequence, determining a risk type of the target object, includes:
Determining a class transformation frequency based on the evaluation classification sequence, determining a risk type of the target object based on the class transformation frequency; and/or the number of the groups of groups,
and determining a class transformation frequency based on the frequency data of the positive class and the negative class in the evaluation classification sequence, and determining the risk type of the target object based on the frequency data of the positive class and the negative class and the class transformation frequency.
7. The method of claim 1, wherein the risk classifying the target object based on the evaluation classification sequence, determining a risk type of the target object, comprises:
and inputting the evaluation classification sequence into a risk discrimination model to obtain the risk type of the target object.
8. A sorting apparatus, comprising:
the text acquisition module is used for acquiring evaluation texts of the target object in different time periods, wherein the different time periods are a plurality of continuous time periods;
the first classification information acquisition module is used for determining first evaluation classification information of the evaluation text based on the evaluation text aiming at the evaluation text in any time period;
A second classification information acquisition module, configured to determine second evaluation classification information of the target object in the time period based on reliability data of the evaluation text and first evaluation classification information of the evaluation text in the time period;
the evaluation classification sequence determining module is used for forming an evaluation classification sequence of the target object based on second evaluation classification information which corresponds to the target object in a plurality of time periods respectively;
and the classification module is used for performing risk classification on the target object based on the evaluation classification sequence and determining the risk type of the target object.
9. An electronic device, the electronic device comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the classification method of any one of claims 1-7.
10. A computer readable storage medium, characterized in that the computer readable storage medium stores computer instructions for causing a processor to implement the classification method according to any one of claims 1-7 when executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311630836.5A CN117633226A (en) | 2023-11-30 | 2023-11-30 | Classification method and device, storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311630836.5A CN117633226A (en) | 2023-11-30 | 2023-11-30 | Classification method and device, storage medium and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117633226A true CN117633226A (en) | 2024-03-01 |
Family
ID=90023000
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311630836.5A Pending CN117633226A (en) | 2023-11-30 | 2023-11-30 | Classification method and device, storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117633226A (en) |
-
2023
- 2023-11-30 CN CN202311630836.5A patent/CN117633226A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114118287A (en) | Sample generation method, sample generation device, electronic device and storage medium | |
| CN112989235A (en) | Knowledge base-based internal link construction method, device, equipment and storage medium | |
| CN114090601B (en) | Data screening method, device, equipment and storage medium | |
| CN112560461A (en) | News clue generation method and device, electronic equipment and storage medium | |
| CN114428677B (en) | Task processing method, processing device, electronic equipment and storage medium | |
| CN113806660B (en) | Data evaluation method, training device, electronic equipment and storage medium | |
| CN114970540A (en) | Method and device for training text audit model | |
| CN114037059A (en) | Pre-training model, model generation method, data processing method and data processing device | |
| CN113051911A (en) | Method, apparatus, device, medium, and program product for extracting sensitive word | |
| CN116228301A (en) | Method, device, equipment and medium for determining target user | |
| CN116226533A (en) | News associated recommendation method, device and medium based on association prediction model | |
| CN113239273B (en) | Method, apparatus, device and storage medium for generating text | |
| CN117633226A (en) | Classification method and device, storage medium and electronic equipment | |
| CN113806541A (en) | Emotion classification method and emotion classification model training method and device | |
| CN113051396A (en) | Document classification identification method and device and electronic equipment | |
| CN117574146B (en) | Text classification labeling method, device, electronic equipment and storage medium | |
| CN116244740B (en) | Log desensitization method and device, electronic equipment and storage medium | |
| CN113377921B (en) | Method, device, electronic equipment and medium for matching information | |
| CN117131197B (en) | Method, device, equipment and storage medium for processing demand category of bidding document | |
| CN115640802A (en) | Evaluation classification method, device, equipment and storage medium for enterprise entities | |
| CN115496594A (en) | Service information processing method, device, equipment and storage medium | |
| CN115146619A (en) | Abnormal short message detection method and device, computer equipment and storage medium | |
| CN116257639A (en) | Logistics knowledge graph generation method, device, equipment and storage medium | |
| CN117290758A (en) | Classification and classification method, device, equipment and medium for unstructured document | |
| CN113360602A (en) | Method, apparatus, device and storage medium for outputting information |
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 |