CN111754352A - Method, device, equipment and storage medium for judging correctness of viewpoint statement - Google Patents
Method, device, equipment and storage medium for judging correctness of viewpoint statement Download PDFInfo
- Publication number
- CN111754352A CN111754352A CN202010574354.2A CN202010574354A CN111754352A CN 111754352 A CN111754352 A CN 111754352A CN 202010574354 A CN202010574354 A CN 202010574354A CN 111754352 A CN111754352 A CN 111754352A
- Authority
- CN
- China
- Prior art keywords
- entity
- viewpoint
- statement
- value
- viewpoint statement
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The invention discloses a method for judging the correctness of viewpoint sentences, which can solve the technical problem that in the prior art, the judgment result is inaccurate because the judgment behavior for judging the correctness of the viewpoint sentences only according to assertions is too unilateral, and comprises the following steps: acquiring a target text; identifying a viewpoint statement in a target text and an entity name, an entity value and an entity time in the viewpoint statement, wherein the viewpoint statement represents that the value of the entity name in the entity time is the entity value; marking out viewpoint sentences in the target text; inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement; and judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement. The invention also discloses a judgment device for the correctness of the viewpoint statement, a computer device and a computer readable storage medium. In addition, the invention also relates to a model training and block chain technology in artificial intelligence.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a method and a device for judging the correctness of viewpoint sentences, computer equipment and a computer-readable storage medium.
Background
It is necessary for investors to find reliable investment analysts before investing, since they want reliable investment analysts to give professional opinions to minimize losses.
The prior art is to evaluate the reliability of investment analysts by analyzing assertions in their view. For example, an investment analyst gives a view of: in 2019, the GDP is estimated to rise by 10%, the prior art extracts the assertion of 'GDP rises by 10%', and then judges the correctness of the viewpoint by analyzing the accuracy of the assertion, so as to evaluate whether an investment analyst is reliable or not according to the correctness of the viewpoint issued by the investment analyst.
However, the inventor researches and discovers that the prior art judges the correctness of the viewpoint only according to the assertion, so that the judgment behavior is too unilateral, and the judgment result is not accurate.
Aiming at the technical problem that in the prior art, the judgment result is inaccurate due to the fact that the judgment behavior of judging the view correctness only according to the assertion is too unilateral, an effective solution is not provided at present.
Disclosure of Invention
The invention aims to provide a method, a system, computer equipment and a computer readable storage medium for judging the correctness of viewpoint sentences, which can solve the technical problem that in the prior art, the judgment result is inaccurate because the judgment behavior of judging the correctness of the viewpoint sentences only according to assertions is too unilateral.
One aspect of the present invention provides a method for judging correctness of an opinion statement, the method comprising: acquiring a target text; identifying a point of view statement in the target text and an entity name, an entity value and an entity time in the point of view statement, the point of view statement characterizing the value of the entity name at the entity time as the entity value; marking out the viewpoint sentences in the target text; inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement; and judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement.
Optionally, the method further comprises: acquiring a plurality of pieces of tense training sample data, wherein each piece of tense training sample data comprises a historical sample marked with a historical viewpoint statement, entity time of the historical viewpoint statement and a tense of the historical viewpoint statement; taking a historical sample with a historical viewpoint statement and entity time of the historical viewpoint statement marked in the acquired temporal training sample data as input of a machine learning algorithm and taking a temporal state of the historical viewpoint statement as output of the machine learning algorithm, and training the machine learning algorithm to obtain a preliminary analysis model; and determining the temporal analysis model according to the preliminary analysis model.
Optionally, determining the temporal analysis model from the preliminary analysis model comprises: acquiring a plurality of pieces of temporal test sample data; inputting the historical sample with the historical viewpoint statement and the entity time of the historical viewpoint statement in the acquired temporal test sample data into the preliminary analysis model to obtain the temporal state of the historical viewpoint statement output by the preliminary analysis model; comparing the output tense of the historical viewpoint statement with the tense of the corresponding historical viewpoint statement in the obtained tense test sample data, and judging whether the accuracy of the primary analysis model is greater than or equal to an accuracy threshold; and when the accuracy of the preliminary analysis model is judged to be more than or equal to the accuracy threshold value, taking the preliminary analysis model as the temporal analysis model.
Optionally, judging whether the view statement is correct according to the entity name, the entity value, the entity time, and the tense of the view statement includes: reading a plurality of pieces of standard data in a target database, wherein each piece of standard data represents a real entity value of any entity name in any entity time; and taking the entity name and the entity time of the viewpoint statement as the entity name and the entity time in the standard data, judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, if so, judging that the viewpoint statement is correct, otherwise, judging that the viewpoint statement is wrong.
Optionally, the determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data includes: when the time state of the viewpoint statement is past, judging whether the entity value of the viewpoint statement is consistent with the entity value in the corresponding standard data, if so, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data; and if the time state of the viewpoint statement is in the future, judging whether the difference range between the entity value of the viewpoint statement and the entity value in the corresponding standard data meets a difference threshold or whether the trend is consistent, if the difference range meets the difference threshold or the trend is consistent, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data.
Optionally, the determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data includes: judging whether the description mode of the entity value of the viewpoint statement is consistent with the description mode of the entity value in the corresponding standard data; and under the condition that the description modes are not consistent, unifying the description modes and then judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data.
Optionally, identifying the opinion statement in the target text and the entity name, entity value, and entity time in the opinion statement comprises: acquiring an identification model, wherein a training set of the identification model comprises a plurality of identification training sample data, and each identification training sample data comprises a historical sample, a historical viewpoint statement in the historical sample and an entity name, an entity value and an entity time of the historical viewpoint statement; inputting the target text into the recognition model, so that the recognition model outputs a viewpoint sentence in the target text and an entity name, an entity value and an entity time in the viewpoint sentence.
Another aspect of the present invention provides an opinion statement correctness judging device, including: the text acquisition module is used for acquiring a target text; the identification data module is used for identifying a viewpoint statement in the target text and an entity name, an entity value and an entity time in the viewpoint statement, wherein the viewpoint statement represents that the value of the entity name at the entity time is the entity value; the annotation statement module is used for annotating the viewpoint statement in the target text; the temporal analysis module is used for inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement; and the viewpoint judging module is used for judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement.
Optionally, the apparatus further comprises: the acquisition sample module is used for acquiring a plurality of pieces of temporal training sample data, wherein each piece of temporal training sample data comprises a historical sample marked with a historical viewpoint statement, and entity time and a temporal state of the historical viewpoint statement; the training model module is used for taking a historical sample with a historical viewpoint statement and entity time of the historical viewpoint statement in the acquired temporal training sample data as input of a machine learning algorithm and taking a temporal state of the historical viewpoint statement as output of the machine learning algorithm, and training the machine learning algorithm to obtain a primary analysis model; and the determining model module is used for determining the temporal analysis model according to the preliminary analysis model.
Optionally, the determining model module is further configured to: acquiring a plurality of pieces of temporal test sample data; inputting the historical sample with the historical viewpoint statement and the entity time of the historical viewpoint statement in the acquired temporal test sample data into the preliminary analysis model to obtain the temporal state of the historical viewpoint statement output by the preliminary analysis model; comparing the output tense of the historical viewpoint statement with the tense of the corresponding historical viewpoint statement in the obtained tense test sample data, and judging whether the accuracy of the primary analysis model is greater than or equal to a preset threshold value; and when the accuracy of the preliminary analysis model is judged to be greater than or equal to the preset threshold value, taking the preliminary analysis model as the temporal analysis model.
Optionally, the viewpoint judging module is further configured to: reading a plurality of pieces of standard data in a target database, wherein each piece of standard data proves the real entity value of any entity name in any entity time; and taking the entity name and the entity time of the viewpoint statement as the entity name and the entity time in the standard data, judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, if so, judging that the viewpoint statement is correct, otherwise, judging that the viewpoint statement is wrong.
Optionally, when determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data, the viewpoint determining module is further configured to: when the time state of the viewpoint statement is past, judging whether the entity value of the viewpoint statement is consistent with the entity value in the corresponding standard data, if so, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data; and if the time state of the viewpoint statement is in the future, judging whether the difference range between the entity value of the viewpoint statement and the entity value in the corresponding standard data meets a preset threshold or whether the trend is consistent, if the difference range meets the difference threshold or the trend is consistent, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data.
Optionally, when determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data, the viewpoint determining module is further configured to: judging whether the description mode of the entity value of the viewpoint statement is consistent with the description mode of the entity value in the corresponding standard data; and under the condition that the description modes are not consistent, unifying the description modes and then judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data.
Optionally, the identification data module is further configured to: acquiring an identification model, wherein a training set of the identification model comprises a plurality of identification training sample data, and each identification training sample data comprises a historical sample, a historical viewpoint statement in the historical sample and an entity name, an entity value and an entity time of the historical viewpoint statement; inputting the target text into the recognition model, so that the recognition model outputs a viewpoint sentence in the target text and an entity name, an entity value and an entity time in the viewpoint sentence.
Yet another aspect of the present invention provides a computer apparatus, comprising: the system comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the method for judging the correctness of the viewpoint statement in any embodiment when executing the computer program.
Yet another aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the method for determining correctness of a viewpoint statement according to any one of the embodiments. Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
After the entity name, the entity value and the entity time of the viewpoint statement in the target text are identified, the temporal state of the viewpoint statement is analyzed by utilizing a pre-trained temporal analysis model based on the entity time of the viewpoint statement and the target text marked with the viewpoint statement, and the correctness of the viewpoint statement is further judged according to the entity name, the entity value, the entity time and the temporal state of the viewpoint statement. Obviously, when judging the correctness of the viewpoint statement, the invention does not only consider the assertion in the viewpoint statement as in the prior art, but further considers the tense of the viewpoint statement, analyzes whether the viewpoint statement is a description of past facts or a prediction of a future form, and further judges the correctness of the viewpoint statement by combining the assertion of the viewpoint statement (i.e. the entity name, the entity value and the entity time of the viewpoint statement) and the tense of the viewpoint statement, thereby achieving the technical effect of improving the accuracy of the judgment result and solving the technical problem that the judgment result is inaccurate because the judgment behavior of judging the correctness of the viewpoint statement only according to the assertion in the prior art is too large.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1 is a flow chart schematically illustrating a method for determining correctness of a viewpoint statement according to an embodiment of the present invention;
FIG. 2 is a diagram schematically illustrating a scheme for determining correctness of a viewpoint statement according to an embodiment of the present invention;
fig. 3 is a block diagram schematically showing a judgment apparatus of the correctness of a viewpoint sentence according to an embodiment of the present invention;
fig. 4 schematically shows a block diagram of a computer apparatus adapted to implement the judgment method of the correctness of a viewpoint statement according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The embodiment of the invention provides a method for judging the correctness of a viewpoint statement, which does not only consider the assertion in the viewpoint statement as in the prior art, but further considers the tense of the viewpoint statement, analyzes whether the viewpoint statement describes past facts or predicts future forms, and further judges the correctness of the viewpoint statement by combining the assertion of the viewpoint statement (namely the entity name, the entity value and the entity time of the viewpoint statement) and the tense of the viewpoint statement, thereby achieving the technical effect of improving the accuracy of the judgment result and solving the technical problem that the judgment result is inaccurate as a result of too much judgment behavior for judging the correctness of the viewpoint statement only according to the assertion in the prior art. Specifically, fig. 1 schematically shows a flowchart of a method for determining correctness of a viewpoint statement according to an embodiment of the present invention, and as shown in fig. 1, the method for determining correctness of a viewpoint statement may include steps S1 to S5, where:
in step S1, a target text is acquired.
The target text can be a document or lecture manuscript published by the user, and the format of the target text can be DOC/DOCX or PDF. If the format of the target text is not a predetermined format which can be recognized by the system, the target text needs to be parsed, information such as words and tables in the target text is recognized, text Recognition (also called chapter Recognition) may include layout Recognition, ORC (Optical Character Recognition) picture Recognition, chapter Recognition, and the like, table Recognition may include table positioning, table parsing, and icon annotation mark, and if the table is in a picture format, the table in the picture format may be converted into a table in a general format by using the ORC technology. Step S2 is further executed to restore the parsed content to the text of the format type recognized by the system as a new target text.
Optionally, the target text may also be classified in advance, such as based on rules or machine learning models, and the classification dimension may include: macro research and report, industry research and report, company research and report, research and report given by research teams and the like, so that the next step of processing can be performed in a targeted manner according to the classified categories. And selecting a recognition model capable of recognizing the text of the category according to the classified category.
Step S2, identifying a viewpoint sentence in the target text and an entity name, an entity value and an entity time in the viewpoint sentence, wherein the viewpoint sentence represents that the value of the entity name is the entity value at the entity time.
The entity name is a subject of judgment in the viewpoint statement, the entity time is a value of when the entity name is expressed in the viewpoint statement and is the entity value, and the entity value can be a specific numerical value or a fuzzy trend or a value range reached by the entity name in the entity time.
For example, the view statements are: the domestic total production value is estimated to increase by 10% in 2019, the entity name of the statement is the domestic total production value, the entity time is 2019, and the entity value is 10%. The opinion statement may include macro economic data.
Alternatively, the embodiment may identify the entity name, the entity value, and the entity time in the viewpoint statement through one model. Specifically, step S2 may include step S21 and step S22, wherein:
step S21, obtaining an identification model, wherein a training set of the identification model comprises a plurality of identification training sample data, and each identification training sample data comprises a historical sample, a historical viewpoint statement in the historical sample and an entity name, an entity value and an entity time of the historical viewpoint statement;
step S22, inputting the target text into the recognition model, so that the recognition model outputs the viewpoint sentence in the target text and the entity name, entity value and entity time in the viewpoint sentence.
In this embodiment, the recognition model may be obtained by training a machine learning algorithm through a training set of the recognition model, wherein during training, a historical sample in the training set of the recognition model is used as an input of the machine learning algorithm, and a historical viewpoint sentence of the historical sample in the training set of the recognition model and an entity name, an entity value, and an entity time of the historical viewpoint sentence are used as an output of the machine learning algorithm, so that the machine learning algorithm may recognize an association rule therein, and form the recognition model capable of recognizing the entity name, the entity value, and the entity time. Further, a recognition model may be applied to the method to recognize the entity name, entity value and entity time of the point of view statement in the target text.
It should be noted that, for different fields, the training set of the recognition model is different, for example, if the recognition model is applicable to the financial field, the training set should include frequently appearing professional terms and proper nouns in the research and report of the financial field, such as economic indicators, market and investment varieties, governments and regulatory agencies, various market participation bodies, companies, countries, regions, leaders, high governments, and the like.
Step S3, marking the opinion statement in the target text.
For example, annotation of a point of view statement may be achieved by tagging the point of view statement in the target text.
Step S4, inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model, so that the temporal analysis model analyzes the temporal of the viewpoint statement.
Tense, the characterising view statement is a prediction of the future or a statement of the past.
For example, if the time when the user gives the viewpoint sentence is 2018, and the viewpoint sentence is estimated that the domestic production total in 2019 increases by 10%, the time of the viewpoint sentence is the future time.
For another example, if the time that the user gives the viewpoint sentence is 2018, and the total domestic production value of the viewpoint sentence is increased by 10% in 2017, the tense of the viewpoint sentence is the past expression.
Obviously, when identifying the tense of the viewpoint sentence, in addition to the entity time in the viewpoint sentence, the time of the user for the viewpoint sentence needs to be considered, and there is no explicit tense grammar rule in chinese, and the expression is flexible, and the phenomenon of word ambiguity is common, so when determining the time of the viewpoint sentence of the user, the determination often needs to be made in combination with the large-span semantics of the context, that is, the time of the user for giving the viewpoint sentence needs to be determined according to the context of the viewpoint sentence in the target text or the time of the target text. Therefore, when analyzing the tense of the viewpoint sentence, it is necessary to input the target text to which the viewpoint sentence is added to the tense analysis model. The temporal analysis model is labeled by a large number of long sentence sequences, is trained and learned by using a machine learning algorithm (such as a deep neural network algorithm), and can accurately identify forecasting, assertions, points of discourse, data of discourse, review, intellectual statement sentences, conditional clauses and the like. Preferably, to further ensure the privacy and security of the temporal analysis model, the temporal analysis model may also be stored in a node of a block chain.
Specifically, the training process of the temporal analysis model may include steps a1 to A3, where:
a1, acquiring a plurality of pieces of tense training sample data, wherein each piece of tense training sample data comprises a historical sample marked with a historical viewpoint statement, entity time of the historical viewpoint statement and a tense of the historical viewpoint statement;
step A2, taking a historical sample with a historical viewpoint statement and entity time of the historical viewpoint statement in the acquired temporal training sample data as input of a machine learning algorithm and taking a temporal state of the historical viewpoint statement as output of the machine learning algorithm, and training the machine learning algorithm to obtain a preliminary analysis model;
step A3, determining the temporal analysis model according to the preliminary analysis model.
In this embodiment, the obtained tense training sample data may be used as a training set to train a machine learning algorithm, so as to obtain a preliminary analysis model, which may also be used to identify the tenses of the viewpoint sentences in the target text, so that the preliminary analysis model may be directly used as a tense analysis model.
However, in order to ensure that the preliminary analysis model can identify the tense of the viewpoint sentence accurately enough, the preliminary analysis model needs to be tested. Specifically, step A3 may include steps a31 to a34, wherein:
a31, acquiring a plurality of pieces of temporal test sample data;
step A32, inputting the historical sample labeled with the historical viewpoint statement in the acquired temporal test sample data and the entity time of the historical viewpoint statement into the preliminary analysis model to obtain the temporal state of the historical viewpoint statement output by the preliminary analysis model;
step A33, comparing the output tense of the historical viewpoint statement with the tense of the corresponding historical viewpoint statement in the acquired tense test sample data, and judging whether the accuracy of the preliminary analysis model is greater than or equal to an accuracy threshold;
and step A34, when the accuracy of the preliminary analysis model is judged to be greater than or equal to the accuracy threshold, taking the preliminary analysis model as the temporal analysis model.
In this embodiment, each piece of temporal test sample data may also include a history sample labeled with a history viewpoint statement, an entity time of the history viewpoint statement, and a temporal state of the history viewpoint statement. When the obtained temporal test sample data is used for testing the preliminary analysis model, aiming at each piece of temporal test sample data, comparing the temporal state of the historical viewpoint statement output by the preliminary analysis model with the temporal state of the historical viewpoint statement in the piece of temporal test sample data, namely comparing the theoretical temporal state of the historical viewpoint statement output by the preliminary analysis model with the real temporal state of the historical viewpoint statement in the corresponding temporal test sample data, if the theoretical temporal state is consistent with the real temporal state, the theoretical temporal state obtained by analyzing the piece of temporal test sample data is correct, otherwise, the theoretical temporal state obtained by analyzing the piece of temporal test sample data is wrong. Further, each piece of acquired temporal test sample data is analyzed and compared, the number of correct temporal test sample data for the theoretical temporal pair of the historical viewpoint statements can be obtained, and then the ratio of the number to the total number of the acquired temporal test sample data is used as the accuracy of the primary analysis model. And then setting a correctness threshold, and when the correctness is greater than the correctness threshold, indicating that the preliminary analysis model can accurately analyze the tense of the viewpoint statement, and at the moment, taking the preliminary analysis model as a tense analysis model to analyze the tense of the target text.
It should be noted that the historical samples included in the temporal training sample data, the historical samples included in the temporal test sample data, and the historical samples included in the training set of the recognition model may be the same or different, and this embodiment does not limit this.
Optionally, in order to speed up the self-learning process of the machine learning algorithm, in this embodiment, some learning rules may be marked in advance for the machine learning algorithm, for example, sentences, words, or values that affect the temporal state of the historical viewpoint sentence are marked in the historical samples as input parameters, so that in the self-learning process of the machine learning algorithm, which sentences, words, or values can be quickly known to determine the temporal state of the historical viewpoint sentence. That is, the input parameters include a history sample that labels the history view and elements that affect the temporal aspect of the history view statement.
Step S5, determining whether the viewpoint statement is correct according to the entity name, the entity value, the entity time, and the tense of the viewpoint statement.
In this embodiment, when the correctness of the viewpoint sentence is evaluated, there are two evaluation schemes, one scheme is that when the entity time of the viewpoint sentence is longer than the current time, the viewpoint sentence is still a future predicted viewpoint compared to the current time, because the future information cannot be accurately obtained at the current time node, the evaluation of the correctness of the viewpoint sentence may adopt a probability value, for example, a viewpoint similar to the viewpoint sentence in other texts is obtained, the viewpoints are all used for predicting the values reached by the entity name of the viewpoint sentence in the target text at the entity time, the values are averaged, if the difference between the entity value of the viewpoint sentence and the average value is within a preset range, the viewpoint sentence is considered to be correct, otherwise, the viewpoint sentence is considered to be incorrect. Another scheme is that when the entity time of the viewpoint statement is less than the current time, the true value of the entity name of the viewpoint statement reached at the entity time is actually present at the current time, so this embodiment may preset a standard library, where a large amount of standard data is stored, and each piece of standard data is used to represent the true entity value of a certain entity name at a certain entity time, and the correctness of the viewpoint statement may be determined by performing a back test on the viewpoint statement. Specifically, step S5 may include steps S51 to S52, in which:
step S51, reading a plurality of pieces of standard data, wherein each piece of standard data represents the real entity value of any entity name in any entity time;
step S52, taking the entity name and the entity time of the viewpoint statement as the entity name and the entity time in the standard data, determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data, if so, determining that the viewpoint statement is correct, otherwise, determining that the viewpoint statement is incorrect.
It should be noted that, since the analyzed temporal state of the viewpoint statement is not based on the current time but based on the time when the user gives the viewpoint statement, even if the entity time of the viewpoint statement is past compared to the current time, as long as the viewpoint statement is used for prediction when the user gives the viewpoint statement, that is, as long as the future of the temporal state of the viewpoint statement is analyzed, in evaluating the correctness of the viewpoint statement, it is only necessary to judge whether the difference range between the entity value of the viewpoint statement and the entity value in the corresponding standard data meets the difference threshold or whether the trend of the entity value of the viewpoint statement meets the trend of the entity value in the corresponding standard data, rather than comparing whether the entity value of the viewpoint statement meets the difference threshold or the trend completely meets the trend of the entity value in the corresponding standard data, the viewpoint statement is considered correct, otherwise the viewpoint statement is considered incorrect. However, when the entity time of the viewpoint statement is less than the time of the viewpoint statement given by the user, that is, the temporal state of the viewpoint statement is past, the real entity value of the entity name in the viewpoint statement at the entity time already exists, and at this time, it is necessary to compare whether the entity value of the viewpoint statement is consistent with the entity value in the corresponding standard data, if so, the viewpoint statement is considered to be correct, otherwise, the viewpoint statement is considered to be incorrect. Specifically, the step S52 of determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data may include steps S521 and S522, where:
step S521, when the temporal state of the viewpoint statement is past, determining whether the entity value of the viewpoint statement is consistent with the entity value in the corresponding standard data, if so, determining that the entity value of the viewpoint statement matches with the entity value in the corresponding standard data, otherwise, determining that the entity value of the viewpoint statement does not match with the entity value in the corresponding standard data;
step S522, when the time state of the viewpoint statement is in the future, determining whether a difference range between the entity value of the viewpoint statement and the entity value in the corresponding standard data meets a difference threshold or a trend is consistent, if the difference range meets the difference threshold or the trend is consistent, determining that the entity value of the viewpoint statement matches the entity value in the corresponding standard data, otherwise determining that the entity value of the viewpoint statement does not match the entity value in the corresponding standard data. That is, if the difference range does not meet the difference threshold or the trend is not consistent, it is determined that the entity value of the viewpoint statement does not match the entity value in the corresponding standard data.
Optionally, the step S52 of determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data may include steps S521 'and S522', in which:
step S521', judging whether the description mode of the entity value of the viewpoint statement is consistent with the description mode of the entity value in the corresponding standard data;
step S522', if the description modes are not consistent, unifying the description modes and then determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data.
The description mode may be a unit, for example, if the entity value in the view statement is 10%, and the entity value in the standard data is 900309 billion, the unit of the entity value in the view statement may be converted into a unit of the entity value in the standard data, for example, a total domestic production value of 2018 is 110%. Of course, the unit of the entity value in the standard data may also be converted into the unit of the entity value in the viewpoint statement, which is not limited in this embodiment. Then, the unified description mode continues to determine whether the two are matched, and the specific matching policy may refer to step S521 and step S522.
After the entity name, the entity value and the entity time of the viewpoint statement in the target text are identified, the temporal state of the viewpoint statement is analyzed by utilizing a pre-trained temporal analysis model based on the entity time of the viewpoint statement and the target text marked with the viewpoint statement, and the correctness of the viewpoint statement is further judged according to the entity name, the entity value, the entity time and the temporal state of the viewpoint statement. Obviously, when judging the correctness of the viewpoint statement, the invention does not only consider the assertion in the viewpoint statement as in the prior art, but further considers the tense of the viewpoint statement, analyzes whether the viewpoint statement is a description of past facts or a prediction of a future form, and further judges the correctness of the viewpoint statement by combining the assertion of the viewpoint statement (i.e. the entity name/entity value and entity time of the viewpoint statement) and the tense of the viewpoint statement, thereby achieving the technical effect of improving the accuracy of the judgment result and solving the technical problem that the judgment result is inaccurate because the judgment behavior of judging the correctness of the viewpoint statement only according to the assertion in the prior art is too large.
Fig. 2 is a schematic diagram schematically illustrating a scheme of judging correctness of a viewpoint statement according to an embodiment of the present invention.
As shown in fig. 2, after obtaining a document (i.e. a target text), a document analysis module is executed first, which mainly includes: the above embodiments may be specifically referred to in the following works, such as file analysis, type analysis, chapter analysis, and icon analysis. Then enter a text understanding module comprising: the method comprises the following steps of Chinese word segmentation, entity recognition, viewpoint extraction and temporal analysis, wherein the Chinese word segmentation is mainly carried out on sentences in a text in a mode of ending word segmentation, the entity recognition and the viewpoint extraction can be completed simultaneously through a recognition model, and the temporal analysis can be completed through a temporal analysis module. Finally, the viewpoint backtesting module is executed, that is, the viewpoint statement correctness is determined, and the viewpoint backtesting module may include: the method comprises the steps of body linking, time alignment, viewpoint matching and utility evaluation, wherein the body linking is used for mapping entity names such as company names, index names or country names in viewpoint sentences to entity names in standard sentences. The time alignment is to map the entity time of the viewpoint statement to the entity time in the standard data, and the standard data for evaluating the correctness of the viewpoint statement can be screened out through the body link and the time alignment. The viewpoint matching is to unify the description modes of the entity values and then compare the entity values in the viewpoint sentences with the entity values of the screened standard data. The utility evaluation sets different judgment rules according to different contents in the viewpoint sentences, such as whether the difference range belongs to a difference threshold value or whether the trends are consistent or not. Further, the number of the users who issue viewpoint sentences and give correct viewpoint sentences can be determined by the above evaluation method, and then the ratio of the number to the total number of the viewpoint sentences issued by the users is used as the criterion for judging the reliability of the users.
Another embodiment of the present invention provides a device for determining correctness of viewpoint statement, which corresponds to the method for determining correctness of viewpoint statement described in the above embodiments, and corresponding technical features and technical effects are not described in detail in this embodiment, and reference may be made to the above embodiments for relevant points. Specifically, fig. 3 schematically shows a block diagram of a viewpoint statement correctness judging device according to an embodiment of the present invention, and as shown in fig. 3, the viewpoint statement correctness judging device 300 may include a text acquiring module 301, an identification data module 302, a labeling statement module 303, a temporal analysis module 304, and a viewpoint judging module 305, where:
an acquiring text module 301, configured to acquire a target text;
an identification data module 302, configured to identify a viewpoint statement in the target text and an entity name, an entity value, and an entity time in the viewpoint statement, where the viewpoint statement represents that a value of the entity name is the entity value at the entity time;
a marking statement module 303, configured to mark the viewpoint statement in the target text;
a temporal analysis module 304, configured to input the entity time in the viewpoint statement and the labeled target text into a temporal analysis model, so that the temporal analysis model analyzes the temporal of the viewpoint statement;
the viewpoint determining module 305 is configured to determine whether the viewpoint statement is correct according to the entity name, the entity value, the entity time, and the temporal state of the viewpoint statement.
Optionally, the apparatus further comprises: the acquisition sample module is used for acquiring a plurality of pieces of temporal training sample data, wherein each piece of temporal training sample data comprises a historical sample marked with a historical viewpoint statement, and entity time and a temporal state of the historical viewpoint statement; the training model module is used for taking a historical sample with a historical viewpoint statement and entity time of the historical viewpoint statement in the acquired temporal training sample data as input of a machine learning algorithm and taking a temporal state of the historical viewpoint statement as output of the machine learning algorithm, and training the machine learning algorithm to obtain a primary analysis model; and the determining model module is used for determining the temporal analysis model according to the preliminary analysis model.
Optionally, the determining model module is further configured to: acquiring a plurality of pieces of temporal test sample data; inputting the historical sample with the historical viewpoint statement and the entity time of the historical viewpoint statement in the acquired temporal test sample data into the preliminary analysis model to obtain the temporal state of the historical viewpoint statement output by the preliminary analysis model; comparing the output tense of the historical viewpoint statement with the tense of the corresponding historical viewpoint statement in the obtained tense test sample data, and judging whether the accuracy of the primary analysis model is greater than or equal to a preset threshold value; and when the accuracy of the preliminary analysis model is judged to be greater than or equal to the preset threshold value, taking the preliminary analysis model as the temporal analysis model.
Optionally, the viewpoint judging module is further configured to: reading a plurality of pieces of standard data in a target database, wherein each piece of standard data proves the real entity value of any entity name in any entity time; and taking the entity name and the entity time of the viewpoint statement as the entity name and the entity time in the standard data, judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, if so, judging that the viewpoint statement is correct, otherwise, judging that the viewpoint statement is wrong.
Optionally, when determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data, the viewpoint determining module is further configured to: when the time state of the viewpoint statement is past, judging whether the entity value of the viewpoint statement is consistent with the entity value in the corresponding standard data, if so, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data; and if the time state of the viewpoint statement is in the future, judging whether the difference range between the entity value of the viewpoint statement and the entity value in the corresponding standard data meets a preset threshold or whether the trend is consistent, if the difference range meets the difference threshold or the trend is consistent, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data.
Optionally, when determining whether the entity value of the viewpoint statement matches the entity value in the corresponding standard data, the viewpoint determining module is further configured to: judging whether the description mode of the entity value of the viewpoint statement is consistent with the description mode of the entity value in the corresponding standard data; and under the condition that the description modes are not consistent, unifying the description modes and then judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data.
Optionally, the identification data module is further configured to: acquiring an identification model, wherein a training set of the identification model comprises a plurality of identification training sample data, and each identification training sample data comprises a historical sample, a historical viewpoint statement in the historical sample and an entity name, an entity value and an entity time of the historical viewpoint statement; inputting the target text into the recognition model, so that the recognition model outputs a viewpoint sentence in the target text and an entity name, an entity value and an entity time in the viewpoint sentence.
Fig. 4 schematically shows a block diagram of a computer apparatus adapted to implement the judgment method of the correctness of a viewpoint statement according to an embodiment of the present invention. In this embodiment, the computer device 400 may be a smart phone, a tablet computer, a notebook computer, a desktop computer, a rack server, a blade server, a tower server, or a rack server (including an independent server or a server cluster composed of a plurality of servers), and the like that execute programs. As shown in fig. 4, the computer device 400 of the present embodiment includes at least, but is not limited to: a memory 401, a processor 402, a network interface 403 communicatively coupled to each other via a system bus. It is noted that FIG. 4 only shows the computer device 400 having components 401 and 403, but it is understood that not all of the shown components are required and that more or fewer components may be implemented instead.
In this embodiment, the memory 403 includes at least one type of computer-readable storage medium, which includes flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, and the like. In some embodiments, the storage 401 may be an internal storage unit of the computer device 400, such as a hard disk or a memory of the computer device 400. In other embodiments, the memory 401 may also be an external storage device of the computer device 400, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. provided on the computer device 400. Of course, the memory 401 may also include both internal and external storage devices for the computer device 400. In the present embodiment, the memory 401 is generally used for storing an operating system and various types of application software installed in the computer device 400, such as program codes of a method for judging correctness of a viewpoint statement, which includes: acquiring a target text; identifying a point of view statement in the target text and an entity name, an entity value and an entity time in the point of view statement, the point of view statement characterizing the value of the entity name at the entity time as the entity value; marking out the viewpoint sentences in the target text; inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement; and judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement.
In this embodiment, the method for determining the correctness of the viewpoint statement stored in the memory 401 can be further divided into one or more program modules and executed by one or more processors (in this embodiment, the processor 402) to complete the present invention.
The network interface 403 may comprise a wireless network interface or a wired network interface, the network interface 403 typically being used to establish communication links between the computer device 400 and other computer devices. For example, the network interface 403 is used to connect the computer apparatus 400 with an external terminal through a network, establish a data transmission channel and a communication link between the computer apparatus 400 and the external terminal, and the like. The network may be a wireless or wired network such as an Intranet (Intranet), the Internet (Internet), a Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), 4G network, 5G network, Bluetooth (Bluetooth), Wi-Fi, etc.
The present embodiment also provides a computer-readable storage medium, including a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, a server, an App application mall, etc., and a computer program is stored thereon, and when executed by a processor, the method for determining correctness of a viewpoint statement includes the steps of: acquiring a target text; identifying a point of view statement in the target text and an entity name, an entity value and an entity time in the point of view statement, the point of view statement characterizing the value of the entity name at the entity time as the entity value; marking out the viewpoint sentences in the target text; inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement; and judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement. Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
It should be noted that the blockchain in the present invention is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
It will be apparent to those skilled in the art that the modules or steps of the embodiments of the invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, embodiments of the invention are not limited to any specific combination of hardware and software.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A method for judging the correctness of an opinion statement, the method comprising:
acquiring a target text;
identifying a point of view statement in the target text and an entity name, an entity value and an entity time in the point of view statement, the point of view statement characterizing the value of the entity name at the entity time as the entity value;
marking out the viewpoint sentences in the target text;
inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement;
and judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement.
2. The method of claim 1, further comprising:
acquiring a plurality of pieces of tense training sample data, wherein each piece of tense training sample data comprises a historical sample marked with a historical viewpoint statement, entity time of the historical viewpoint statement and a tense of the historical viewpoint statement;
taking a historical sample with a historical viewpoint statement and entity time of the historical viewpoint statement marked in the acquired temporal training sample data as input of a machine learning algorithm and taking a temporal state of the historical viewpoint statement as output of the machine learning algorithm, and training the machine learning algorithm to obtain a preliminary analysis model;
and determining the temporal analysis model according to the preliminary analysis model.
3. The method of claim 2, wherein determining the temporal analysis model from the preliminary analysis model comprises:
acquiring a plurality of pieces of temporal test sample data;
inputting the historical sample with the historical viewpoint statement and the entity time of the historical viewpoint statement in the acquired temporal test sample data into the preliminary analysis model to obtain the temporal state of the historical viewpoint statement output by the preliminary analysis model;
comparing the output tense of the historical viewpoint statement with the tense of the corresponding historical viewpoint statement in the obtained tense test sample data, and judging whether the accuracy of the primary analysis model is greater than or equal to an accuracy threshold;
and when the accuracy of the preliminary analysis model is judged to be more than or equal to the accuracy threshold value, taking the preliminary analysis model as the temporal analysis model.
4. The method of claim 1, wherein determining whether the point of view statement is correct based on the entity name, the entity value, the entity time, and the temporal state of the point of view statement comprises:
reading a plurality of pieces of standard data, wherein each piece of standard data represents a real entity value of any entity name at any entity time;
and taking the entity name and the entity time of the viewpoint statement as the entity name and the entity time in the standard data, judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, if so, judging that the viewpoint statement is correct, otherwise, judging that the viewpoint statement is wrong.
5. The method of claim 4, wherein determining whether the entity value of the point of view statement matches an entity value in the corresponding criterion data comprises:
when the time state of the viewpoint statement is past, judging whether the entity value of the viewpoint statement is consistent with the entity value in the corresponding standard data, if so, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data;
and if the time state of the viewpoint statement is in the future, judging whether the difference range between the entity value of the viewpoint statement and the entity value in the corresponding standard data meets a difference threshold or whether the trend is consistent, if the difference range meets the difference threshold or the trend is consistent, judging that the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data, otherwise, judging that the entity value of the viewpoint statement is not matched with the entity value in the corresponding standard data.
6. The method of claim 4, wherein determining whether the entity value of the point of view statement matches an entity value in the corresponding criterion data comprises:
judging whether the description mode of the entity value of the viewpoint statement is consistent with the description mode of the entity value in the corresponding standard data;
and under the condition that the description modes are not consistent, unifying the description modes and then judging whether the entity value of the viewpoint statement is matched with the entity value in the corresponding standard data.
7. The method of claim 1, wherein identifying the opinion statement in the target text and the entity name, entity value, and entity time in the opinion statement comprises:
acquiring an identification model, wherein a training set of the identification model comprises a plurality of identification training sample data, and each identification training sample data comprises a historical sample, a historical viewpoint statement in the historical sample and an entity name, an entity value and an entity time of the historical viewpoint statement;
inputting the target text into the recognition model, so that the recognition model outputs a viewpoint sentence in the target text and an entity name, an entity value and an entity time in the viewpoint sentence.
8. An apparatus for judging correctness of a viewpoint sentence, the apparatus comprising:
the text acquisition module is used for acquiring a target text;
the identification data module is used for identifying a viewpoint statement in the target text and an entity name, an entity value and an entity time in the viewpoint statement, wherein the viewpoint statement represents that the value of the entity name at the entity time is the entity value;
the annotation statement module is used for annotating the viewpoint statement in the target text;
the temporal analysis module is used for inputting the entity time in the viewpoint statement and the labeled target text into a temporal analysis model so that the temporal analysis model analyzes the temporal of the viewpoint statement;
and the viewpoint judging module is used for judging whether the viewpoint statement is correct or not according to the entity name, the entity value, the entity time and the tense of the viewpoint statement.
9. A computer system comprising a plurality of computer devices, each computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processors of the plurality of computer devices collectively implement the method of any one of claims 1 to 7 when the computer program is executed by the processors.
10. A computer-readable storage medium comprising a plurality of storage media, each storage medium having a computer program stored thereon, wherein the computer programs stored by the plurality of storage media, when executed by a processor, collectively implement the method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010574354.2A CN111754352A (en) | 2020-06-22 | 2020-06-22 | Method, device, equipment and storage medium for judging correctness of viewpoint statement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010574354.2A CN111754352A (en) | 2020-06-22 | 2020-06-22 | Method, device, equipment and storage medium for judging correctness of viewpoint statement |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111754352A true CN111754352A (en) | 2020-10-09 |
Family
ID=72675694
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010574354.2A Pending CN111754352A (en) | 2020-06-22 | 2020-06-22 | Method, device, equipment and storage medium for judging correctness of viewpoint statement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111754352A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112395870A (en) * | 2020-11-30 | 2021-02-23 | 联想(北京)有限公司 | Text processing method and equipment |
| CN113220885A (en) * | 2021-05-21 | 2021-08-06 | 支付宝(杭州)信息技术有限公司 | Text processing method and system |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556580A (en) * | 2009-05-20 | 2009-10-14 | 北京工商大学 | Stock comment classification system based on analysis of discourse structure and method |
| US20150012331A1 (en) * | 2013-07-05 | 2015-01-08 | 1st Call Consulting, Pte Ltd. | Computer-Implemented Intelligence Tool |
| CN108764981A (en) * | 2018-05-16 | 2018-11-06 | 平安科技（深圳）有限公司 | Server, the processing method of text data and storage medium |
| CN109299252A (en) * | 2018-08-17 | 2019-02-01 | 北京奇虎科技有限公司 | The viewpoint polarity classification method and device of stock comment based on machine learning |
| CN109766540A (en) * | 2018-12-10 | 2019-05-17 | 平安科技（深圳）有限公司 | Generic text information extracting method, device, computer equipment and storage medium |
| CN110209786A (en) * | 2019-05-28 | 2019-09-06 | 北京百度网讯科技有限公司 | It is display methods, device, computer equipment and the storage medium of non-class answer |
| CN110263319A (en) * | 2019-03-21 | 2019-09-20 | 国家计算机网络与信息安全管理中心 | A kind of scholar's viewpoint abstracting method based on web page text |
| CN110378516A (en) * | 2019-06-17 | 2019-10-25 | 平安科技（深圳）有限公司 | Analyst's portrait generation method, device, equipment and computer readable storage medium |
| CN110738046A (en) * | 2018-07-03 | 2020-01-31 | 百度在线网络技术（北京）有限公司 | Viewpoint extraction method and device |
| DE202020102105U1 (en) * | 2020-04-16 | 2020-04-29 | Robert Bosch Gmbh | Device for the automated generation of a knowledge graph |
-
2020
- 2020-06-22 CN CN202010574354.2A patent/CN111754352A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556580A (en) * | 2009-05-20 | 2009-10-14 | 北京工商大学 | Stock comment classification system based on analysis of discourse structure and method |
| US20150012331A1 (en) * | 2013-07-05 | 2015-01-08 | 1st Call Consulting, Pte Ltd. | Computer-Implemented Intelligence Tool |
| CN108764981A (en) * | 2018-05-16 | 2018-11-06 | 平安科技（深圳）有限公司 | Server, the processing method of text data and storage medium |
| CN110738046A (en) * | 2018-07-03 | 2020-01-31 | 百度在线网络技术（北京）有限公司 | Viewpoint extraction method and device |
| CN109299252A (en) * | 2018-08-17 | 2019-02-01 | 北京奇虎科技有限公司 | The viewpoint polarity classification method and device of stock comment based on machine learning |
| CN109766540A (en) * | 2018-12-10 | 2019-05-17 | 平安科技（深圳）有限公司 | Generic text information extracting method, device, computer equipment and storage medium |
| WO2020119075A1 (en) * | 2018-12-10 | 2020-06-18 | 平安科技（深圳）有限公司 | General text information extraction method and apparatus, computer device and storage medium |
| CN110263319A (en) * | 2019-03-21 | 2019-09-20 | 国家计算机网络与信息安全管理中心 | A kind of scholar's viewpoint abstracting method based on web page text |
| CN110209786A (en) * | 2019-05-28 | 2019-09-06 | 北京百度网讯科技有限公司 | It is display methods, device, computer equipment and the storage medium of non-class answer |
| CN110378516A (en) * | 2019-06-17 | 2019-10-25 | 平安科技（深圳）有限公司 | Analyst's portrait generation method, device, equipment and computer readable storage medium |
| DE202020102105U1 (en) * | 2020-04-16 | 2020-04-29 | Robert Bosch Gmbh | Device for the automated generation of a knowledge graph |
Non-Patent Citations (2)
| Title |
|---|
| JASMINA SMAILOVIĆ ET AL.: "Stream-based Active Learning for Sentiment Analysis in the Financial Domain", 《INFORMATION SCIENCES》, vol. 285, pages 181 - 203 * |
| 兰秋军 等: "融合句法信息的金融论坛文本情感计算研究", 《现代图书情报技术》, no. 4, pages 64 - 71 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112395870A (en) * | 2020-11-30 | 2021-02-23 | 联想(北京)有限公司 | Text processing method and equipment |
| CN113220885A (en) * | 2021-05-21 | 2021-08-06 | 支付宝(杭州)信息技术有限公司 | Text processing method and system |
| CN113220885B (en) * | 2021-05-21 | 2023-02-07 | 支付宝(杭州)信息技术有限公司 | Text processing method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109992664B (en) | Dispute focus label classification method and device, computer equipment and storage medium | |
| CN112270196B (en) | Entity relationship identification method and device and electronic equipment | |
| CN107102993B (en) | User appeal analysis method and device | |
| CN112860841B (en) | Text emotion analysis method, device, equipment and storage medium | |
| CN111191275A (en) | Sensitive data identification method, system and device | |
| CN108170468B (en) | Method and system for automatically detecting annotation and code consistency | |
| CN111783016B (en) | Website classification method, device and equipment | |
| CN112052684A (en) | Named entity identification method, device, equipment and storage medium for power metering | |
| CN111984792A (en) | Website classification method and device, computer equipment and storage medium | |
| CN112416778A (en) | Test case recommendation method and device and electronic equipment | |
| CN111460250A (en) | Image data cleaning method, image data cleaning device, image data cleaning medium, and electronic apparatus | |
| CN112347226A (en) | Document knowledge extraction method and device, computer equipment and readable storage medium | |
| CN112052305A (en) | Information extraction method and device, computer equipment and readable storage medium | |
| CN111754352A (en) | Method, device, equipment and storage medium for judging correctness of viewpoint statement | |
| CN113728321A (en) | Using a set of training tables to accurately predict errors in various tables | |
| CN111190946A (en) | Report generation method and device, computer equipment and storage medium | |
| CN113312899B (en) | Text classification method and device and electronic equipment | |
| CN110781673B (en) | Document acceptance method and device, computer equipment and storage medium | |
| CN116010688A (en) | User behavior label identification method, system, equipment and storage medium | |
| CN113591459B (en) | Address standardization processing method and device, electronic equipment and readable storage medium | |
| CN113822040A (en) | Subjective question marking and scoring method and device, computer equipment and storage medium | |
| CN111290953B (en) | Method and device for analyzing test logs | |
| CN110941713B (en) | Self-optimizing financial information block classification method based on topic model | |
| CN110705258A (en) | Text entity identification method and device | |
| CN114969334B (en) | Abnormal log detection method and device, electronic equipment and readable storage 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 |