CN113553423B - Scenario information processing method and device, electronic equipment and storage medium - Google Patents
Scenario information processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113553423B CN113553423B CN202110757681.6A CN202110757681A CN113553423B CN 113553423 B CN113553423 B CN 113553423B CN 202110757681 A CN202110757681 A CN 202110757681A CN 113553423 B CN113553423 B CN 113553423B
- Authority
- CN
- China
- Prior art keywords
- scenario
- session
- highlight
- score
- outlier
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/338—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3346—Query execution using probabilistic model
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Artificial Intelligence (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to a scenario information processing method, a scenario information processing device, electronic equipment and a storage medium, wherein the scenario information processing method comprises the following steps: acquiring scenario text corresponding to each session in the scenario file; according to the scenario text corresponding to each scenario, calculating an analysis index corresponding to each scenario, wherein the analysis index is used for analyzing the highlight information of each scenario in the scenario file; for each session, calculating a highlight score for each session based on an analysis index corresponding to the session; a highlight variation curve for characterizing the highlight variation of each session is generated based on the highlight scores of each session. According to the embodiment of the application, the analysis indexes are calculated according to the scenario text corresponding to each scenario, so that the characteristic information in the scenario text is quantized, the subsequent calculation of the highlight score based on the analysis indexes is facilitated, the automatic generation of the highlight change curve based on the highlight score is facilitated, and the generation efficiency and accuracy of the highlight change curve are improved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a scenario information processing method, apparatus, electronic device, and storage medium.
Background
The scenario mainly consists of a speech line and stage instructions. Dialogs, monologs, and bystandings are all speech substitutes, and are often represented by singing words in drama and opera. Stage instructions in the script are descriptive written in the breath of the drama author. The method comprises the steps of crossing the time and place of occurrence of the scenario, describing the image characteristics, the body actions and the internal activities of the characters in the scenario, describing the scenes and the atmosphere, and requiring the aspects of scenery, light, sound effects and the like.
Unlike novels or film-like video viewing curves, the highlights of the scenes cannot be measured by user behaviors and can only be scored based on text content, so that a score maker is required to have profound understanding on the business characteristics of the scenes, meanwhile, the existing text understanding work results are utilized to draw the scene-level highlights change curves, however, the efficiency of manually drawn highlights change curves is low, and curves drawn by different score makers may be greatly different and have poor accuracy.
Disclosure of Invention
In order to solve the technical problems or at least partially solve the technical problems, the application provides a scenario information processing method, a scenario information processing device, electronic equipment and a storage medium.
In a first aspect, the present application provides a scenario information processing method, including:
acquiring scenario text corresponding to each session in the scenario file;
according to the scenario text corresponding to each scenario, calculating an analysis index corresponding to each scenario, wherein the analysis index is used for analyzing the highlight information of each scenario in the scenario file;
for each session, calculating a highlight score for each session based on an analysis index corresponding to the session;
a highlight variation curve for characterizing the highlight variation of each session is generated based on the highlight scores of each session.
Optionally, the analysis index includes: one or more of important role attendance ratio, number of attendance roles, total play of the attendance roles, interaction times of core characters, total emotion intensity of the attendance roles and characteristic values of the plot points;
the important role departure ratio is the proportion of the number of important roles of the scene departure to the total number of the roles of the scene departure, and the important roles are a plurality of roles with the largest play in the scenario file;
the number of the out-of-the-field roles is the total number of out-of-the-field roles in the session;
the total play of the out-of-the-field roles is the sum of the plays of each role in the session;
the interaction times of the core characters are the interaction times among the core characters in the scene, and the core characters are a plurality of characters of the character relation core in the script file;
the total emotion intensity of the character appearing in the field is the sum of emotion values of the character appearing in the field;
the characteristic value of the plot point is the importance degree of the plot in the plot point to which the plot point belongs.
Optionally, calculating an analysis index corresponding to each session according to the scenario text corresponding to each session, including:
acquiring characters of the scene, the play data of each character and the total number of the characters of the scene in the script file, wherein the play data is determined according to the behaviors of the characters in the scene and the occurrence times of the dialogs;
sorting the play data of each role in the script file, and selecting a plurality of roles with top ranking of the play to obtain a plurality of important roles in the script file;
counting the number of roles matched with any important role in the roles of the scene departure to obtain a first number of important roles of the scene departure;
and calculating the ratio between the first quantity and the total quantity to obtain the important role departure ratio.
Optionally, calculating an analysis index corresponding to each session according to the scenario text corresponding to each session, including:
acquiring scenario node information of scenario nodes to which the scenario belongs, the scenario included in the scenario node information and the arrangement sequence of the scenario;
and calculating the characteristic value of the scenario node according to the field ranking order and the position of the field in the field ranking order.
Optionally, for each session, calculating a highlight score of each session based on the analysis index corresponding to the session, including:
converting each analysis index into a characteristic score according to the analysis index corresponding to each occasion;
and multiplying the characteristic score of each session by a corresponding weight coefficient to obtain the highlight score of each session.
Optionally, generating a highlight variation curve for characterizing highlight variation conditions of each session based on the highlight score of each session includes:
performing exponential smoothing on the highlight score of each occasion to obtain a first intermediate score of each occasion;
performing outlier detection in the first intermediate score of each session to obtain a head outlier set, a tail outlier set and a normal point set;
normalizing the head outlier set to a first interval, normalizing the tail outlier set to a second interval, and normalizing the normal point set to a third interval to obtain a second intermediate score of each session, wherein a smaller boundary threshold of the first interval is greater than or equal to a larger boundary threshold of the third interval, and a smaller boundary threshold of the third interval is greater than or equal to a larger boundary threshold of the second interval;
a highlight variation curve is generated based on the second intermediate score for each session.
Optionally, performing outlier detection in the first intermediate score of each session to obtain a head outlier set, a tail outlier set and a normal point set, including:
determining a first intermediate score greater than a first preset threshold as a head outlier, and constructing a head outlier set containing the head outlier;
determining a first intermediate score smaller than a second preset threshold as a tail outlier, and constructing a tail outlier set containing the tail outlier, wherein the first preset threshold is larger than the second preset threshold;
the first intermediate scores other than the head outlier and the tail outlier among the first intermediate scores of all the shots are determined as normal points, and a normal point set including normal points is constructed.
In a second aspect, the present application provides a scenario information processing apparatus, comprising:
the acquisition module is used for acquiring the scenario text corresponding to each occasion in the scenario file;
the first calculation module is used for calculating an analysis index corresponding to each scenario according to the scenario text corresponding to each scenario, wherein the analysis index is used for analyzing the highlight information of each scenario in the scenario file;
the second calculation module is used for calculating the highlight score of each occasion based on the analysis index corresponding to each occasion;
and the generation module is used for generating a highlight change curve for representing the highlight change condition of each session based on the highlight score of each session.
In a third aspect, the present application provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete communication with each other through the communication bus;
a memory for storing a computer program;
and the processor is used for realizing the scenario information processing method according to any one of the first aspect when executing the program stored in the memory.
In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon a program of scenario information processing method, which when executed by a processor, implements the steps of any one of the scenario information processing methods described in the first aspect.
Compared with the prior art, the technical scheme provided by the embodiment of the application has the following advantages:
according to the embodiment of the application, the scenario text corresponding to each occasion in the scenario file is firstly obtained, then the analysis index corresponding to each occasion is calculated according to the scenario text corresponding to each occasion, then the highlight score of each occasion is calculated according to the analysis index corresponding to each occasion, and finally the highlight change curve for representing the highlight change condition of each occasion can be generated based on the highlight score of each occasion.
According to the embodiment of the application, the analysis indexes are calculated according to the scenario text corresponding to each scenario, so that the characteristic information in the scenario text is quantized, the subsequent calculation of the highlight score based on the analysis indexes is facilitated, the automatic generation of the highlight change curve based on the highlight score is facilitated, and the generation efficiency and accuracy of the highlight change curve are improved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.
In order to more clearly illustrate the embodiments of the application or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, and it will be obvious to a person skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Fig. 1 is a flowchart of a scenario information processing method provided in an embodiment of the present application;
fig. 2 is a block diagram of a scenario information processing apparatus according to an embodiment of the present application;
fig. 3 is a block diagram of an electronic device according to an embodiment of the present application;
fig. 4 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Unlike novels or film-like video viewing curves, the highlights of the scenes cannot be measured by user behaviors and can only be scored based on text content, so that a score maker is required to have profound understanding on the business characteristics of the scenes, meanwhile, the existing text understanding work results are utilized to draw the scene-level highlights change curves, however, the efficiency of manually drawn highlights change curves is low, and curves drawn by different score makers may be greatly different and have poor accuracy. Therefore, the embodiment of the application provides a scenario information processing method, a scenario information processing device, electronic equipment and a storage medium, wherein the scenario information processing method can be applied to a computer.
As shown in fig. 1, the scenario information processing method may include the steps of:
step S101, acquiring a scenario text corresponding to each session in a scenario file;
in the embodiment of the application, the scenario file comprises a plurality of occasions, and the text content corresponding to each occasion is scenario text. Each scenario text may record a play character, interactions between characters (e.g., behavioral interactions, dialogue interactions, etc.), emotional states of each character, etc.
Step S102, calculating an analysis index corresponding to each scenario according to scenario text corresponding to each scenario, wherein the analysis index is used for analyzing the highlight information of each scenario in the scenario file;
in practical application, the judgment of the script essence chroma is generally carried out by the following method:
1) Generally, group play (more people are out) and the number of times the important character is out are more wonderful according to the abundance of the appearance of the character and whether the important character is out or not;
2) According to the number of the interaction times of the characters, the more the interaction between the characters is, the more the content is wonderful;
3) Action and contrast occupy more shots, content is more colorful, descriptive language is more shots, and content is relatively poor;
4) The number of times of extreme value occurrence of the emotion of the character and the emotion intensity can be used as the basis for judging the climax.
5) The middle part of a complete episode (consisting of multiple episodes) is often more wonderful, the highlight should be lower, with the head and tail following other episodes.
Based on this, in the embodiment of the present application, an analysis index is set, where the analysis index includes: one or more of important role attendance ratio, number of attendance roles, total play of the attendance roles, interaction times of core characters, total emotion intensity of the attendance roles and characteristic values of the plot points;
the important role departure ratio is the proportion of the number of important roles of the scene departure to the total number of the roles of the scene departure, the important roles are a plurality of roles with the largest play in the scenario file, and the calculation mode of the important role departure ratio is described in detail below;
the number of the out-of-field roles is the total number of out-of-field roles in the session, namely, for each session, the total number of out-of-field roles in the session is counted, so that the number v2 of out-of-field roles can be obtained, v2=role_count, role_count represents the number of out-of-field roles in the session, the number of out-of-field roles can represent whether group play occurs in the session or not, and the occurrence group play represents a more highlight;
the total play of the out-of-the-field roles is the sum of the plays of each role in the session, namely, for each session, the play of each role in the session can be counted for each role, and the play of each role is added to obtain the sum of the plays of each role in the session. The total play of the out-of-field roles is used for expressing the behaviors, the occurrence times and the like of all out-of-field roles in the scene.
In the embodiment of the application, the play can refer to the performance workload born by the character in the session, and the number of the play can be obtained by counting the occurrence times of the behavior interaction, the dialogue interaction, the single emotion description and the like in the session by way of example;
the total play v3 of the out-of-the-field character can be calculated by the following formula:
where drama_count represents the play of character i in the session, and P represents the session's set of outgoing characters.
The number of interactions between core characters is the number of interactions between core characters in the scene, the core characters are a plurality of characters of a character relationship core marked in advance in a script file, namely, for each scene, the core characters appearing in the scene can be determined according to a core character list in a script related file, the number of interactions between the core characters (the interactions can comprise behavioral interactions, dialogue interactions and the like) is counted, in the embodiment of the application, whether preset interactive verbs exist in the script content of the scene can be detected, if the interactive verbs are detected, whether at least two character nouns or character significands exist in a preset character length range (such as left and right 5 characters and the like) near the interactive verbs can be detected, if the at least two character nouns or character significands exist, the detection of the interactions can be determined, and the exemplary interactive verbs comprise: beat, pull, push, listen, say, etc., for example: the script content is as follows: "I beat the shoulder with a small bright" and since "I" and "small bright" are detected near "beat", then one interaction can be determined to be detected; for another example: "I am to small, since" I am "and" small "are detected near" say ", it is also possible to determine that an interaction is detected; for another example: "I have unzipped", since only "I" a personal noun or person reference is detected near "pull", it can be determined that no interaction is detected.
The interaction times v4 of the core character can be calculated by the following formula:
wherein, the interactive_count represents the number of interactions of the core character (top 10) in the field.
The total emotion intensity (namely seven emotion intensity) of the characters appearing in the scene is the sum of emotion values of the characters appearing in the scene, namely, for each scene, the emotion value of each character can be calculated according to a preset algorithm for calculating the emotion value of the character appearing in the scene, and the emotion values of the characters appearing in the scene are added to obtain the total emotion intensity;
the total emotional intensity v5 of the out-of-the-field character can be calculated by the following formula:
in the embodiment of the application, whether preset emotion words exist in a preset character length range (such as left and right 3 characters and the like) near a target character in the transcript content of the scene or not can be detected, each preset emotion word has a corresponding emotion score, the emotion score corresponding to each emotion word is searched based on each detected emotion word, and emotion scores corresponding to all emotion words of the target character are accumulated to obtain the emotion value of the target character.
For example: the emotion words include: smile, laugh, cry, 24696 cry, etc., with emotion score for "laugh" being 2 points, emotion score for laugh being 4 points, emotion score for cry being 3 points, emotion score for cry being 5 points, assuming: the script content is as follows: "light red smile" says, since "smile" is detected near the target character "light red", it can be determined that the emotion value is plus 2 points; for another example: "laugh with reddish shade" since "laugh" is detected near the target character "reddish shade", the emotion value can be determined plus 4 points; for another example: "reddish 24696" cry "since" 24696 "is detected in the vicinity of the target character" reddish ", the emotion value can be determined plus 5 minutes, etc.
The characteristic value of a plot is the importance of the scene in the node of the plot, the plot is a particular event or event in the drama terminology of a movie or a television show, the event is tightly woven into the story and the story is turned to another direction, and one plot may contain at least one scene, and the characteristic value of the plot is calculated in a manner described in detail below.
In the step, information such as the play roles and interactions among the roles (such as behavioral interactions and dialogue interactions) and the emotional states of each role in the scenario text corresponding to each session can be extracted, important role information, core role information, interaction information among each role and other roles and the like are extracted from the scenario related file, and analysis indexes corresponding to each session are calculated based on the extracted information.
Step S103, calculating the highlight score of each session based on the analysis index corresponding to each session;
since the measurement difference of each analysis index is large and the influence of each analysis index on the highlight degree of the scenario text is different, in this step, the analysis indexes of each scene can be converted into the same measurement, and a weight is added to each analysis index, so as to obtain the highlight degree score of each scene.
Step S104, generating a highlight degree change curve for representing the highlight degree change condition of each session based on the highlight degree score of each session.
In this step, a highlight degree score for each session may be set to the Y axis, and a highlight degree change curve may be generated using the session as the X axis.
According to the embodiment of the application, the scenario text corresponding to each occasion in the scenario file is firstly obtained, then the analysis index corresponding to each occasion is calculated according to the scenario text corresponding to each occasion, then the highlight score of each occasion is calculated according to the analysis index corresponding to each occasion, and finally the highlight change curve for representing the highlight change condition of each occasion can be generated based on the highlight score of each occasion.
According to the embodiment of the application, the analysis indexes are calculated according to the scenario text corresponding to each scenario, so that the characteristic information in the scenario text is quantized, the subsequent calculation of the highlight score based on the analysis indexes is facilitated, the automatic generation of the highlight change curve based on the highlight score is facilitated, and the generation efficiency and accuracy of the highlight change curve are improved.
In yet another embodiment of the present application, step S102 calculates an analysis index corresponding to each session according to scenario text corresponding to each session, including:
step 201, obtaining the character of the scene, the play data of each character and the total number of characters of the scene in the script file.
In the embodiment of the application, the play data is determined according to the behavior of the character in the scene and the occurrence times of the dialect;
step 202, sorting the play data of each character in the script file, and selecting a plurality of characters with top ranking of the play, so as to obtain a plurality of important characters in the script file;
in the step, the play data of each character in the script file can be ordered according to the number of the plays, and a plurality of important characters ranked at the top in the ordering are selected. Illustratively, the characters in the transcript file may be ordered in order of more than less than one, or the characters in the transcript file may be ordered in order of less than more than one.
Step 203, counting the number of roles matched with any important role in the roles of the scene departure, and obtaining a first number of important roles of the scene departure;
and step 204, calculating the ratio between the first quantity and the total quantity to obtain the important role departure ratio.
In the embodiment of the application, the important role departure ratio can be calculated through the following formula:
v_1=top_role/role_count
where top_role refers to the first number of important roles in the departure of the scene, and role_count refers to the total number of roles in the departure of the scene.
In the embodiment of the application, the number of important roles with the largest play in the scenario file is calculated as follows:
where N represents the total number of out-of-house characters in the script file.
According to the embodiment of the application, the analysis index of the important role departure ratio can be automatically calculated, the quantification of the characteristic information in the scenario text is realized, the subsequent calculation of the highlight score based on the analysis index is facilitated, the automatic generation of the highlight change curve based on the highlight score is facilitated, and the generation efficiency and the accuracy of the highlight change curve are improved.
In yet another embodiment of the present application, calculating an analysis index corresponding to each session according to scenario text corresponding to each session includes:
step 303, acquiring scenario node information of a scenario node to which the scenario belongs, the scenario contained in the scenario node information and the arrangement sequence of the scenario;
and step 304, calculating the characteristic value of the scenario node according to the field ranking order and the position of the field in the field ranking order.
The plot point feature value v6 may be calculated by:
v 6 =f(plot_num)
where plot_num is the field identity, exemplary, f () function represents: setting the characteristic values of the scenario nodes at the two ends as 0.5 according to the sequence of the scenario nodes, and sequentially adding 0.5 to the middle scenario as the characteristic value of the scenario node of the corresponding scenario, wherein the scenario characteristic value of the scenario node is 0.
For example: if 7 scenes belong to the same emotion node, the emotion node characteristic values of the 7 scenes are [0.5,1,1.5,2,1.5,1,0.5 ]).
According to the embodiment of the application, the analysis index of the characteristic value of the plot point can be automatically calculated, the characteristic information in the plot text is quantized, the subsequent calculation of the highlight score based on the analysis index is facilitated, the automatic generation of the highlight change curve based on the highlight score is facilitated, and the generation efficiency and accuracy of the highlight change curve are improved.
In still another embodiment of the present application, for each session, calculating a highlight score for each session based on an analysis index corresponding to the session, includes:
step 401, converting each analysis index into a feature score according to the analysis index corresponding to each scene;
in practical application, for each analysis index, a set F of values corresponding to the analysis index in n fields is constructed 1 ={v 11 ,v 12 ,…,v 1n -a }; each analysis index can then be converted into a feature score using the following Z-core normalization method:
the mean () function and the std () function respectively represent the mean value and the standard deviation, and other analysis indexes sequentially adopt the mode to convert the analysis indexes into feature scores.
Step 402, multiplying the feature score of each session by the corresponding weight coefficient to obtain the highlight score of each session.
The formula for calculating the precision score is as follows:
wherein t is the field, i represents the ith feature, w i =[0.6,1.0,0.8,1.0,1.2,0.5]And (5) scoring corresponding weight coefficients for each characteristic.
Based on the above equation, a set of highlight scores s= { S for each session can be calculated 1 ,s 2 ,…,s n S1 is the highlight score for the first shot, s2 is the highlight score for the second shot, and sn is the highlight score for the nth shot.
The embodiment of the application can automatically calculate the highlight score of each occasion, is convenient for automatically generating the highlight change curve based on the highlight score, and improves the generation efficiency and accuracy of the highlight change curve.
In yet another embodiment of the present application, generating a highlight variation curve for characterizing highlight variation conditions for each session based on highlight scores for each session includes:
step 501, performing exponential smoothing on the highlight score of each session to obtain a first intermediate score of each session;
the highlight score for each session may be exponentially smoothed as follows
s i =α×s i +(1-α)×s i-1 ,i∈{2,3,…,n}
Wherein a is a preset natural number.
Step 502, performing outlier detection in the first middle score of each session to obtain a head outlier set, a tail outlier set and a normal point set;
in this step, a first intermediate score that is greater than a first preset threshold may be determined as head outliers and a set of head outliers including the head outliers is constructed; determining a first intermediate score smaller than a second preset threshold as a tail outlier, and constructing a tail outlier set containing the tail outlier, wherein the first preset threshold is larger than the second preset threshold; the first intermediate scores other than the head outlier and the tail outlier among the first intermediate scores of all the shots are determined as normal points, and a normal point set including normal points is constructed.
Illustratively, assuming q1 and q3 are 0.25 and 0.75 minutes of s, respectively, for each si, if si < q1-3× (q 3-q 1), si is defined as the head outlier, if si > q3+3× (q 3-q 1), si is defined as the tail outlier, and the rest of the points are normal points. Wherein, the quantile means that a group of digits are arranged from small to large, the digits at the position of 1/4 are 0.25 quantiles, and the digits at the position of 3/4 are 0.75 quantiles.
Step 503, normalizing the head outlier set to a first interval, normalizing the tail outlier set to a second interval, and normalizing the normal point set to a third interval, so as to obtain a second intermediate score of each session.
Wherein the smaller boundary threshold of the first section is greater than or equal to the larger boundary threshold of the third section, which is greater than or equal to the larger boundary threshold of the second section;
because the data point value is positive or negative, and the obtained numerical range of each scenario has larger difference, normalization is between 0 and 100, and the normalization is more standard, for example: there is a set of data: [ -100, -99, -1,0,1,2,3,2,1,3,99,100 ], the set of numbers is calculated by outlier method, -100, -99 is the tail outlier, 99 and 100 are the head outliers (i.e. the difference from the overall data is too large, directly plotting the curves results in the data of these individual outliers to flatten the middle curve and show no fluctuations), thus normalizing tail outliers [ -100, -99] between 3-7, head outliers [99,100] between 93-100, and normal points [ -1,0,1,2,3,2,1,3] between 11-85.
In this step, the head outlier set Sh, the normal point set Sn, and the tail outlier set St may be normalized according to the following formulas:
wherein h is min =3,h max =7,n min =11,n max =85,t min =93,t max =100。
Step 504, generating a highlight variation curve based on the second intermediate score of each session.
The embodiment of the application can automatically carry out exponential smoothing processing, outlier detection and normalization processing on the highlight score of each occasion, and is convenient for smoothing the generated highlight change curve.
In order to facilitate understanding, the application also provides an embodiment in practical application, as shown in fig. 2, assuming that the scenario file includes n occasions, namely, 1,2 and … … occasions, namely, n, the number of important role plays, the number of roles of the plays, total plays of roles of the roles, the interaction times of the core characters, total emotion intensity of the roles and characteristic values of the nodes can be calculated, then each analysis index is converted into characteristic scores to obtain v1, v2 and … … v6, weighted average is carried out on v1, v2 and … … v6 to obtain the highlight score of each occasion, exponential smoothing, outlier detection and normalization processing are carried out on the highlight score of each occasion, and finally, a highlight change curve is generated based on the processed data.
In still another embodiment of the present application, as shown in fig. 3, there is also provided a scenario information-processing apparatus including:
the acquiring module 11 is configured to acquire scenario text corresponding to each session in the scenario file;
a first calculation module 12, configured to calculate an analysis index corresponding to each scenario according to scenario text corresponding to each scenario, where the analysis index is used to analyze highlight information of each scenario in the scenario file;
a second calculation module 13 for calculating, for each session, a highlight score for each session based on an analysis index corresponding to the session;
a generation module 14 for generating a highlight variation curve for characterizing the highlight variation situation of each session based on the highlight score of each session.
In yet another embodiment of the present application, there is provided an electronic device including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;
a memory for storing a computer program;
and the processor is used for realizing the scenario information processing method according to any method embodiment when executing the program stored in the memory.
According to the electronic equipment provided by the embodiment of the application, the processor firstly acquires the scenario text corresponding to each occasion in the scenario file by executing the program stored in the memory, then calculates the analysis index corresponding to each occasion according to the scenario text corresponding to each occasion, calculates the highlight score of each occasion according to the analysis index corresponding to each occasion, and finally generates the highlight change curve for representing the highlight change condition of each occasion based on the highlight score of each occasion.
According to the embodiment of the application, the analysis indexes are calculated according to the scenario text corresponding to each scenario, so that the characteristic information in the scenario text is quantized, the subsequent calculation of the highlight score based on the analysis indexes is facilitated, the automatic generation of the highlight change curve based on the highlight score is facilitated, and the generation efficiency and accuracy of the highlight change curve are improved.
The communication bus 1140 mentioned above for the electronic device may be a Peripheral Component Interconnect (PCI) bus or an Extended Industrial Standard Architecture (EISA) bus, etc. The communication bus 1140 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in fig. 4, but not only one bus or one type of bus.
The communication interface 1120 is used for communication between the electronic device and other devices described above.
The memory 1130 may include Random Access Memory (RAM) or non-volatile memory (non-volatile memory), such as at least one magnetic disk memory. Optionally, the memory may also be at least one memory device located remotely from the aforementioned processor.
The processor 1110 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSP), application Specific Integrated Circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
In still another embodiment of the present application, there is also provided a computer-readable storage medium having stored thereon a program of scenario information processing method, which when executed by a processor, implements the steps of the scenario information processing method described in any one of the method embodiments described above.
It should be noted that in this document, relational terms such as "first" and "second" and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, 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 one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing is only a specific embodiment of the application to enable those skilled in the art to understand or practice the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A scenario information processing method, comprising:
acquiring scenario text corresponding to each session in the scenario file;
according to the scenario text corresponding to each scenario, calculating an analysis index corresponding to each scenario, wherein the analysis index is used for analyzing the highlight information of each scenario in the scenario file;
for each session, calculating a highlight score for each session based on an analysis index corresponding to the session;
for each session, calculating a highlight score for each session based on the analysis index corresponding to the session, comprising:
converting each analysis index into a characteristic score according to the analysis index corresponding to each occasion;
multiplying the characteristic score of each session by a corresponding weight coefficient to obtain a highlight score of each session;
generating a highlight variation curve for representing the highlight variation condition of each session based on the highlight score of each session;
generating a highlight variation curve for characterizing highlight variation conditions of each session based on the highlight scores of each session, comprising:
performing exponential smoothing on the highlight score of each occasion to obtain a first intermediate score of each occasion;
performing outlier detection in the first intermediate score of each session to obtain a head outlier set, a tail outlier set and a normal point set;
normalizing the head outlier set to a first interval, normalizing the tail outlier set to a second interval, and normalizing the normal point set to a third interval to obtain a second intermediate score of each session, wherein a smaller boundary threshold of the first interval is greater than or equal to a larger boundary threshold of the third interval, and a smaller boundary threshold of the third interval is greater than or equal to a larger boundary threshold of the second interval;
a highlight variation curve is generated based on the second intermediate score for each session.
2. The scenario information processing method according to claim 1, wherein the analysis index comprises: one or more of important role attendance ratio, number of attendance roles, total play of the attendance roles, interaction times of core characters, total emotion intensity of the attendance roles and characteristic values of the plot points;
the important role departure ratio is the proportion of the number of important roles of the scene departure to the total number of the roles of the scene departure, and the important roles are a plurality of roles with the largest play in the scenario file;
the number of the out-of-the-field roles is the total number of out-of-the-field roles in the session;
the total play of the out-of-the-field roles is the sum of the plays of each role in the session;
the interaction times of the core characters are the interaction times among the core characters in the scene, and the core characters are a plurality of characters of the character relation core in the script file;
the total emotion intensity of the character appearing in the field is the sum of emotion values of the character appearing in the field;
the characteristic value of the plot point is the importance degree of the plot in the plot point to which the plot point belongs.
3. The scenario information processing method according to claim 2, wherein calculating an analysis index corresponding to each session from scenario text corresponding to each session comprises:
acquiring characters of the scene, the play data of each character and the total number of the characters of the scene in the script file, wherein the play data is determined according to the behaviors of the characters in the scene and the occurrence times of the dialogs;
sorting the play data of each role in the script file, and selecting a plurality of roles with top ranking of the play to obtain a plurality of important roles in the script file;
counting the number of roles matched with any important role in the roles of the scene departure to obtain a first number of important roles of the scene departure;
and calculating the ratio between the first quantity and the total quantity to obtain the important role departure ratio.
4. The scenario information processing method according to claim 2, wherein calculating an analysis index corresponding to each session from scenario text corresponding to each session comprises:
acquiring scenario node information of scenario nodes to which the scenario belongs, the scenario included in the scenario node information and the arrangement sequence of the scenario;
and calculating the characteristic value of the scenario node according to the field ranking order and the position of the field in the field ranking order.
5. The scenario information processing method according to claim 1, wherein performing outlier detection in the first intermediate score of each session to obtain a head outlier set, a tail outlier set, and a normal point set, comprises:
determining a first intermediate score greater than a first preset threshold as a head outlier, and constructing a head outlier set containing the head outlier;
determining a first intermediate score smaller than a second preset threshold as a tail outlier, and constructing a tail outlier set containing the tail outlier, wherein the first preset threshold is larger than the second preset threshold;
the first intermediate scores other than the head outlier and the tail outlier among the first intermediate scores of all the shots are determined as normal points, and a normal point set including normal points is constructed.
6. A scenario information processing apparatus, comprising:
the acquisition module is used for acquiring the scenario text corresponding to each occasion in the scenario file;
the first calculation module is used for calculating an analysis index corresponding to each scenario according to the scenario text corresponding to each scenario, wherein the analysis index is used for analyzing the highlight information of each scenario in the scenario file;
the second calculation module is used for calculating the highlight score of each occasion based on the analysis index corresponding to each occasion; for each session, calculating a highlight score for each session based on the analysis index corresponding to the session, comprising: converting each analysis index into a characteristic score according to the analysis index corresponding to each occasion; multiplying the characteristic score of each session by a corresponding weight coefficient to obtain a highlight score of each session;
the generating module is used for generating a highlight change curve for representing the highlight change condition of each session based on the highlight score of each session; generating a highlight variation curve for characterizing highlight variation conditions of each session based on the highlight scores of each session, comprising: performing exponential smoothing on the highlight score of each occasion to obtain a first intermediate score of each occasion; performing outlier detection in the first intermediate score of each session to obtain a head outlier set, a tail outlier set and a normal point set; normalizing the head outlier set to a first interval, normalizing the tail outlier set to a second interval, and normalizing the normal point set to a third interval to obtain a second intermediate score of each session, wherein a smaller boundary threshold of the first interval is greater than or equal to a larger boundary threshold of the third interval, and a smaller boundary threshold of the third interval is greater than or equal to a larger boundary threshold of the second interval; a highlight variation curve is generated based on the second intermediate score for each session.
7. The electronic equipment is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are communicated with each other through the communication bus;
a memory for storing a computer program;
and the processor is used for realizing the scenario information processing method according to any one of claims 1-5 when executing the program stored in the memory.
8. A computer-readable storage medium, wherein a program of a scenario information processing method is stored on the computer-readable storage medium, the program of the scenario information processing method implementing the steps of the scenario information processing method according to any one of claims 1 to 5 when executed by a processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110757681.6A CN113553423B (en) | 2021-07-05 | 2021-07-05 | Scenario information processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110757681.6A CN113553423B (en) | 2021-07-05 | 2021-07-05 | Scenario information processing method and device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113553423A CN113553423A (en) | 2021-10-26 |
CN113553423B true CN113553423B (en) | 2023-10-10 |
Family
ID=78102717
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110757681.6A Active CN113553423B (en) | 2021-07-05 | 2021-07-05 | Scenario information processing method and device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113553423B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114065742B (en) * | 2021-11-19 | 2023-08-25 | 马上消费金融股份有限公司 | Text detection method and device |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101585830B1 (en) * | 2015-06-22 | 2016-01-15 | 이호석 | Storytelling system and method according to emotion of audience |
CN107193969A (en) * | 2017-05-25 | 2017-09-22 | 南京大学 | A kind of method for automatically generating novel text emotion curve and predicting recommendation |
CN107480136A (en) * | 2017-08-02 | 2017-12-15 | 陈雷 | It is a kind of to be applied to the method that feeling curve is analyzed in screen play |
CN108563688A (en) * | 2018-03-15 | 2018-09-21 | 西安影视数据评估中心有限公司 | A kind of movie and television play principle thread recognition methods |
CN110457691A (en) * | 2019-07-26 | 2019-11-15 | 北京影谱科技股份有限公司 | Feeling curve analysis method and device based on drama role |
CN110458428A (en) * | 2019-07-26 | 2019-11-15 | 北京小土科技有限公司 | A kind of excellent metrization assessment system of screen play |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8644971B2 (en) * | 2009-11-09 | 2014-02-04 | Phil Weinstein | System and method for providing music based on a mood |
US20160117311A1 (en) * | 2014-10-22 | 2016-04-28 | Thomson Licensing | Method and Device for Performing Story Analysis |
US20170300748A1 (en) * | 2015-04-02 | 2017-10-19 | Scripthop Llc | Screenplay content analysis engine and method |
-
2021
- 2021-07-05 CN CN202110757681.6A patent/CN113553423B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101585830B1 (en) * | 2015-06-22 | 2016-01-15 | 이호석 | Storytelling system and method according to emotion of audience |
CN107193969A (en) * | 2017-05-25 | 2017-09-22 | 南京大学 | A kind of method for automatically generating novel text emotion curve and predicting recommendation |
CN107480136A (en) * | 2017-08-02 | 2017-12-15 | 陈雷 | It is a kind of to be applied to the method that feeling curve is analyzed in screen play |
CN108563688A (en) * | 2018-03-15 | 2018-09-21 | 西安影视数据评估中心有限公司 | A kind of movie and television play principle thread recognition methods |
CN110457691A (en) * | 2019-07-26 | 2019-11-15 | 北京影谱科技股份有限公司 | Feeling curve analysis method and device based on drama role |
CN110458428A (en) * | 2019-07-26 | 2019-11-15 | 北京小土科技有限公司 | A kind of excellent metrization assessment system of screen play |
Also Published As
Publication number | Publication date |
---|---|
CN113553423A (en) | 2021-10-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108073568B (en) | Keyword extraction method and device | |
Yu et al. | Video paragraph captioning using hierarchical recurrent neural networks | |
CN107818798B (en) | Customer service quality evaluation method, device, equipment and storage medium | |
CN110267119B (en) | Video precision and chroma evaluation method and related equipment | |
CN110928994A (en) | Similar case retrieval method, similar case retrieval device and electronic equipment | |
CN112183334B (en) | Video depth relation analysis method based on multi-mode feature fusion | |
WO2021114841A1 (en) | User report generating method and terminal device | |
CN111723784B (en) | Risk video identification method and device and electronic equipment | |
CN110263854B (en) | Live broadcast label determining method, device and storage medium | |
CN111538827A (en) | Case recommendation method and device based on content and graph neural network and storage medium | |
CN110210540B (en) | Cross-social media user identity recognition method and system based on attention mechanism | |
CN114612839B (en) | Short video analysis processing method, system and computer storage medium | |
CN112995690B (en) | Live content category identification method, device, electronic equipment and readable storage medium | |
CN113553423B (en) | Scenario information processing method and device, electronic equipment and storage medium | |
CN112041809A (en) | Automatic addition of sound effects to audio files | |
CN113298015A (en) | Video character social relationship graph generation method based on graph convolution network | |
CN116567351B (en) | Video processing method, device, equipment and medium | |
CN115525740A (en) | Method and device for generating dialogue response sentence, electronic equipment and storage medium | |
CN111626058B (en) | Based on CR 2 Image-text double-coding realization method and system of neural network | |
CN114548263A (en) | Method and device for verifying labeled data, computer equipment and storage medium | |
KR20200081613A (en) | Apparatus and method for detecting increase section of conflict in narrative | |
CN111507084A (en) | Statement difficulty determining method and system | |
CN116523024B (en) | Training method, device, equipment and storage medium of recall model | |
CN116758450B (en) | Video question-answering method based on collaborative attention reasoning of space-time diagram and similarity diagram | |
JP5054653B2 (en) | Viewing impression estimation method and apparatus, program, and computer-readable recording medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |