Project task popularity assessment method and device

Abstract

The invention provides a project task popularity assessment method and device, and the method specifically comprises the following steps: (1) splitting a project task text into a multi-task description text, and extracting task keywords and feature vectors of all tasks; (2) carrying out orthogonalization processing on the feature vectors of the tasks; (3) performing intensity standardization on the feature vector after orthogonalization processing; and (4) predicting the activeness of the corresponding task based on the orthogonalized feature vector and the intensity-standardized feature vector. According to the method, the activeness of the task is predicted according to the dynamic characteristics of the task, the progress change of the task can be responded in real time, and the adaptability and the flexibility in multiple types of projects are improved.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, and more specifically, this invention relates to a method and apparatus for evaluating the popularity of project tasks. Background Technology

[0002] As project management becomes increasingly complex, enterprises face the core challenge of dynamically adjusting multi-task collaboration. In fields such as IT, manufacturing, and construction engineering, project schedules often deviate due to unclear task dependencies and non-linear influencing factors.

[0003] Project popularity is an important indicator for measuring the level of attention a project receives, the intensity of resource investment, and the activity level of its progress. It directly affects task priority settings, resource allocation strategies, and the frequency of progress monitoring. High-popularity projects are usually characterized by high urgency, large resource consumption, and multiple parallel tasks, requiring more frequent and refined progress tracking and deviation response mechanisms. Low-popularity projects, on the other hand, may require less resource investment and have stable progress, so the frequency of deviation detection and correction can be appropriately relaxed.

[0004] The determination of project popularity usually relies on historical data statistics, such as the number of tasks completed and the rate of resource consumption; subjective evaluation, such as regular scoring by project managers or teams; or simple threshold rules, such as setting indicators such as the number of days of task delay and the percentage of budget overrun.

[0005] The above methods have the following shortcomings: (1) They are highly subjective and lack consistency and objective quantitative basis; (2) They are difficult to adapt to multiple types and stages of projects and have weak generalization ability. Summary of the Invention

[0006] In view of this, this application provides a method for evaluating the popularity of project tasks, which aims to improve at least one of the above-mentioned problems.

[0007] Specifically, the following technical solutions are included:

[0008] On the one hand, this application provides a method for evaluating the popularity of project tasks, the method being as follows:

[0009] (1) Split the project task text into multi-task description text, and extract the task keywords and feature vectors of each task;

[0010] (2) Orthogonalize the feature vectors of the task;

[0011] (3) Standardize the intensity of the eigenvectors after orthogonalization;

[0012] (4) Based on the orthogonalized feature vectors and intensity-normalized feature vectors, predict the activity of the corresponding task.

[0013] In some embodiments of the present invention, task keywords are... Input the pre-trained word embedding model Extract task keywords eigenvectors eigenvectors It is expressed as follows:

[0014] ;

[0015] in, Indicates task keywords eigenvectors.

[0016] In some embodiments of the present invention, the orthogonalization process of the task's feature vectors is as follows:

[0017] Set a set of orthogonal bases according to the task requirements of each task. The feature vectors of the task are orthogonalized based on orthogonal bases, and the specific formula is as follows:

[0018] ;

[0019] in, Let represent the feature vector of the i-th task keyword. Let j-th task keyword be an orthogonal basis. For feature vectors The eigenvectors after orthogonal processing.

[0020] In some embodiments of the present invention, feature vectors The strength standardization process is as follows:

[0021] ;

[0022] in, For feature vectors The intensity-normalized feature vector These are configurable parameters. Let be the orthogonal basis of the task corresponding to the i-th task keyword.

[0023] In some embodiments of the present invention, the method for determining the activity level of a task is as follows:

[0024] (41) Extract the feature vectors of each task keyword and eigenvectors Input a Poisson observation model, and the Poisson observation model will output the monthly delivery volume for the corresponding task. ;

[0025] (42) Detect monthly delivery volume Is it greater than If the detection result is yes, the corresponding task is considered to be highly active; if the detection result is no, step (43) is executed.

[0026] (43) The feature vector of the task keywords and eigenvectors Input the Bernoulli observation model, and the Bernoulli observation model will output the monthly delivery volume for the corresponding task. ;

[0027] (44) Detect monthly delivery volume Is it less than If the detection result is yes, the corresponding task is identified as having low activity; if the detection result is no, proceed to step (45).

[0028] (45) Calculate the monthly delivery volume and First difference and monthly delivery volume and The second difference is used to determine the activity level. If the first difference is less than the second difference, the corresponding task is considered to have high activity. If the first difference is greater than the second difference, the corresponding task is considered to have low activity.

[0029] In some embodiments of the present invention, the Poisson observation model Specifically as follows:

[0030] ;

[0031] in, , This represents the eigenvectors after orthogonal processing. The feature vector after intensity normalization. Representing the eigenvector With feature vectors The intensity of the inner product projection. These are adjustment parameters used to adjust the observation sensitivity. Let be the dimension of the feature vector corresponding to the task. For strength parameters, This indicates the monthly delivery volume for the corresponding task.

[0032] In some embodiments of the present invention, the Bernoulli observation model Specifically as follows:

[0033] ;

[0034] in, Indicates the dose attenuation factor. , Representing the eigenvector With feature vectors The intensity of the inner product projection. This indicates the monthly delivery volume for the corresponding task.

[0035] On the other hand, embodiments of this application provide a project task popularity assessment device, the device comprising:

[0036] The task feature extraction module, orthogonal processing module, intensity standardization module, and activity prediction module are connected sequentially.

[0037] The task feature extraction module is used to split the project task text into multi-task description text and extract the task keywords and feature vectors of each task.

[0038] The orthogonal processing module is used to orthogonalize the feature vectors of the task.

[0039] The intensity normalization module is used to normalize the intensity of the orthogonalized feature vectors.

[0040] The activity prediction module is used to predict the activity level of the corresponding task based on the orthogonalized feature vectors and the intensity-normalized feature vectors.

[0041] In some embodiments of the present invention, the activity prediction module includes: a Poisson observation model and a Bernoulli observation model connected to the orthogonal processing module and the intensity standardization module, and a prediction unit connected to the Poisson observation model and the Bernoulli observation model.

[0042] The orthogonal processing module and the intensity normalization module respectively process the feature vectors of each task. eigenvectors Input a Poisson observation model, and the Poisson observation model will output the monthly delivery volume for the corresponding task. Predicting the monthly delivery volume of the unit. Is it greater than If the detection result is yes, the corresponding task is considered to have high activity; if the detection result is no, the orthogonal processing module and the intensity standardization module respectively process the feature vectors of each task. eigenvectors Input the Bernoulli observation model, and the Bernoulli observation model will output the monthly delivery volume for the corresponding task. Predicting monthly delivery volume of unit inspection Is it less than If the detection result is yes, the corresponding task is considered to have low activity; if the detection result is no, the monthly delivery volume is calculated. and First difference and monthly delivery volume and The second difference is used to determine the activity level. If the first difference is less than the second difference, the corresponding task is considered to have high activity. If the first difference is greater than the second difference, the corresponding task is considered to have low activity.

[0043] In some embodiments of the present invention, the orthogonal processing module performs orthogonalization processing on the feature vectors of the task, and the specific processing procedure is as follows:

[0044] Set a set of orthogonal bases according to the task requirements of each task. The feature vectors of the task are orthogonalized based on orthogonal bases, and the specific formula is as follows:

[0045] ;

[0046] in, Let represent the feature vector of the i-th task keyword. Let j-th task keyword be an orthogonal basis. For feature vectors The eigenvectors after orthogonal processing.

[0047] This invention predicts task activity based on the dynamic characteristics of the task, enabling real-time response to changes in task progress and improving adaptability and flexibility in various types of projects. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A flowchart of the project task popularity assessment method provided in this embodiment of the invention;

[0050] Figure 2 This is a schematic diagram of the structure of the project task heat assessment device according to an embodiment of the present invention;

[0051] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0052] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0053] Unless otherwise defined, all technical terms used in the embodiments of this application have the same meaning as commonly understood by one of ordinary skill in the art.

[0054] This invention first splits the project task text into multi-task description text, then extracts the task keywords and feature vectors for each task; next, it predicts task activity based on the task feature vectors, wherein... Figure 1 The flowchart of the project task popularity assessment method provided in this embodiment of the invention is as follows:

[0055] (1) Split the project task text into multi-task description text, and extract the task keywords and feature vectors of each task;

[0056] The semantic analysis and feature extraction of project task text includes: preprocessing the project task text to eliminate noise, and generating task feature vectors based on semantic analysis; and orthogonalizing the task feature vectors to eliminate semantic correlation between tasks.

[0057] First, the project task text is split into multiple task description texts, and task keywords are extracted from each task description text. Add to corpus Chinese corpus ,in, Short text snippets, such as "Proficient in Python data analysis," represent task keywords; similarly, task description texts like "Proficient in Python data analysis" or "Familiar with cloud computing architecture design" extract several task keywords from the task description text. .

[0058] Then, keywords for each task were analyzed. Generates that follow a standard normal distribution A random vector, where, for An identity matrix of order 1 ensures the orthogonality of different task vectors;

[0059] Task keywords Input the pre-trained word embedding model (Word2Vec or GloVe) to extract task keywords. eigenvectors eigenvectors It is expressed as follows:

[0060] (1)

[0061] in, Indicates task keywords eigenvectors.

[0062] (2) Orthogonalize the task feature vectors to ensure semantic independence among the task vectors; adjust the feature vectors using the Gram-Schmidt orthogonalization algorithm. The direction is to eliminate semantic relevance interference, thereby achieving semantic correction. Specifically, this process involves using pre-trained word vectors... Set up a set of orthogonal bases according to task requirements. The task feature vectors are orthogonalized based on orthogonal bases, and the specific formula is as follows:

[0063] (2)

[0064] in, Let represent the feature vector of the i-th task keyword. Let j-th task keyword be an orthogonal basis. Represents the projection coefficient, numerator The denominator represents the inner product (projection) of the eigenvectors and the orthogonal basis. The squared magnitude of the orthogonal basis vectors (normalization factor) is derived from the eigenvectors. Subtracting the projection components in all existing orthogonal directions yields a new eigenvector perpendicular to the existing subspace. .

[0065] (3) Standardize the intensity of the eigenvectors after orthogonalization;

[0066] Intensity standardization: for new feature vectors Scale the module length to meet the constraints. ;

[0067] (3)

[0068] in, The feature vector is scaled by its modulus (intensity normalization). This is a configurable parameter with a default value. By adjusting parameters Control the level of discrimination in task descriptions; Adjust the magnitude of the unit vector from 1 to ; This represents normalization, transforming orthogonalized vectors into unit vectors (magnitude 1). Its purpose is to retain only direction information and eliminate bias caused by differences in the length of the original vectors. Constraints Ensure task independence (e.g., "Python programming" [0.7, -0.3, 1.2] and "cloud computing" [0.5, 1.1, -0.8] are orthogonal, these vectors maintain orientation independence in space, avoiding semantic overlap;

[0069] For example, in high-activity scenarios (such as the IT project phase): Set parameters It can accurately match core tasks (such as "distributed system design"), but the same recommendation results decrease by 23%, while the click-through rate of top projects increases by 41%; while in long-tail scenarios (such as traditional manufacturing): It can expand the scope of task association (such as associating "mechanical drawing" with "CAD operation"), but the exposure of long-tail projects increases by 67%.

[0070] (4) Based on the orthogonalized feature vectors and intensity-normalized feature vectors, predict the activity of the corresponding task.

[0071] In scenarios with a high volume of task submissions (high dose), the Poisson observation model is used to observe the activity level of the corresponding tasks; in scenarios with fewer projects (low dose), the Bernoulli observation model is used to observe the activity level of the corresponding tasks. Based on this, the specific method for determining task activity level is as follows:

[0072] (41) Extract the feature vectors of each task keyword and eigenvectors Input a Poisson observation model, and the Poisson observation model will output the monthly delivery volume for the corresponding task. ;

[0073] (42) Detect monthly delivery volume Is it greater than If the detection result is yes, the corresponding task is considered to be highly active; if the detection result is no, step (43) is executed.

[0074] (43) The feature vector of the task keywords and eigenvectors Input the Bernoulli observation model, and the Bernoulli observation model will output the monthly delivery volume for the corresponding task. ;

[0075] (44) Detect monthly delivery volume Is it less than If the detection result is yes, the corresponding task is identified as having low activity; if the detection result is no, proceed to step (45).

[0076] (45) Calculate the monthly delivery volume and First difference and monthly delivery volume and The second difference is used to determine the activity level. If the first difference is less than the second difference, the corresponding task is considered to have high activity. If the first difference is greater than the second difference, the corresponding task is considered to have low activity.

[0077] In this embodiment of the invention, a Poisson observation model is applicable to high-activity tasks (such as software development positions). Specifically as follows:

[0078] (4)

[0079] in, , This represents the eigenvectors after orthogonal processing. The feature vector after intensity normalization. Representing the eigenvector With feature vectors The intensity of the inner product projection. These are adjustment parameters used to adjust the observation sensitivity. Let be the dimension of the feature vector corresponding to the task. For strength parameters, This indicates the monthly delivery volume for the corresponding task.

[0080] In this embodiment of the invention, a Bernoulli observation model is applicable to low-activity tasks (such as blockchain architects). Specifically as follows:

[0081] (5)

[0082] in, This represents the dose attenuation factor, which maps the degree of continuous matching to binary event probabilities using an exponential function. , This indicates the monthly delivery volume for the corresponding task.

[0083] This invention predicts task activity based on the dynamic characteristics of the task, enabling real-time response to changes in task progress and improving adaptability and flexibility in various types of projects.

[0084] Figure 2 This is a schematic diagram of the project task popularity assessment device according to an embodiment of the present invention. For ease of explanation, only the parts related to the embodiment of the present invention are shown. The device includes:

[0085] The task feature extraction module, orthogonal processing module, intensity standardization module, and activity prediction module are connected sequentially.

[0086] The task feature extraction module is used to split the project task text into multi-task description text and extract the task keywords and feature vectors of each task.

[0087] The orthogonal processing module is used to orthogonalize the feature vectors of the task.

[0088] The intensity normalization module is used to normalize the intensity of the orthogonalized feature vectors.

[0089] The activity prediction module is used to predict the activity level of the corresponding task based on the orthogonalized feature vectors and the intensity-normalized feature vectors.

[0090] The task feature extraction module splits the project task text into multi-task description text, extracts task keywords from the task description text, and then... Input the pre-trained word embedding model Extract task keywords eigenvectors eigenvectors It is expressed as follows:

[0091] ;

[0092] in, Indicates task keywords eigenvectors.

[0093] The orthogonal processing module performs orthogonalization of the task's feature vectors, and the specific processing procedure is as follows:

[0094] Set a set of orthogonal bases according to the task requirements of each task. The feature vectors of the task are orthogonalized based on orthogonal bases, and the specific formula is as follows:

[0095] ;

[0096] in, Let i represent the feature vector of the i-th task keyword. Let j-th task keyword be an orthogonal basis. For feature vectors The eigenvectors after orthogonal processing.

[0097] The intensity normalization module is used to normalize the orthogonalized eigenvectors. Strength standardization is performed, and the specific process is as follows:

[0098] ;

[0099] in, For feature vectors The intensity-normalized feature vector These are configurable parameters. Let be the orthogonal basis of the task corresponding to the i-th task keyword.

[0100] In this embodiment of the invention, the activity prediction module includes: a Poisson observation model and a Bernoulli observation model, both connected to an orthogonal processing module and an intensity standardization module; and a prediction unit connected to the Poisson observation model and the Bernoulli observation model.

[0101] The orthogonal processing module and the intensity normalization module respectively process the feature vectors of each task. eigenvectors Input a Poisson observation model, and the Poisson observation model will output the monthly delivery volume for the corresponding task. Predicting the monthly delivery volume of the unit. Is it greater than If the detection result is yes, the corresponding task is considered to have high activity; if the detection result is no, the orthogonal processing module and the intensity standardization module respectively process the feature vectors of each task. eigenvectors Input the Bernoulli observation model, and the Bernoulli observation model will output the monthly delivery volume for the corresponding task. Predicting monthly delivery volume of unit inspection Is it less than If the detection result is yes, the corresponding task is considered to have low activity; if the detection result is no, the monthly delivery volume is calculated. and First difference and monthly delivery volume and The second difference is used to determine the activity level. If the first difference is less than the second difference, the corresponding task is considered to have high activity. If the first difference is greater than the second difference, the corresponding task is considered to have low activity.

[0102] In this embodiment of the invention, the Poisson observation model Specifically as follows:

[0103] ;

[0104] in, , This represents the eigenvectors after orthogonal processing. The feature vector after intensity normalization. Calculate eigenvectors With feature vectors The intensity of the inner product projection. These are adjustment parameters used to adjust the observation sensitivity. Let be the dimension of the feature vector corresponding to the task. For strength parameters, This indicates the monthly delivery volume for the corresponding task.

[0105] In this embodiment of the invention, the Bernoulli observation model Specifically as follows:

[0106] ;

[0107] in, Indicates the dose attenuation factor. , This indicates the monthly delivery volume for the corresponding task.

[0108] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only.

[0109] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for evaluating the popularity of project tasks, characterized in that, The method is as follows: (1) Split the project task text into multi-task description text, and extract the task keywords and feature vectors of each task; (2) Orthogonalize the feature vectors of the task; (3) Standardize the intensity of the eigenvectors after orthogonalization; (4) Based on the orthogonalized feature vectors and intensity-normalized feature vectors, predict the activity of the corresponding task.

2. The project task popularity assessment method as described in claim 1, characterized in that, Task keywords Input the pre-trained word embedding model Extract task keywords eigenvectors eigenvectors It is expressed as follows: ； in, Indicates task keywords eigenvectors.

3. The project task popularity assessment method as described in claim 1, characterized in that, The orthogonalization process of the task's feature vectors is as follows: Set a set of orthogonal bases according to the task requirements of each task. The feature vectors of the task are orthogonalized based on orthogonal bases, and the specific formula is as follows: ； in, Let represent the feature vector of the i-th task keyword. Let j-th task keyword be an orthogonal basis. For feature vectors The eigenvectors after orthogonal processing.

4. The project task popularity assessment method as described in claim 1, characterized in that, Feature vector The strength standardization process is as follows: ； in, For feature vectors The intensity-normalized feature vector These are configurable parameters. Let be the orthogonal basis of the task corresponding to the i-th task keyword.

5. The project task popularity assessment method as described in claim 1, characterized in that, The method for determining task activity is as follows: (41) Extract the feature vectors of each task keyword and eigenvectors Input a Poisson observation model, and the Poisson observation model will output the monthly delivery volume for the corresponding task. ; (42) Detect monthly delivery volume Is it greater than If the detection result is yes, the corresponding task is considered to be highly active; if the detection result is no, step (43) is executed. (43) The feature vector of the task keywords and eigenvectors Input the Bernoulli observation model, and the Bernoulli observation model will output the monthly delivery volume for the corresponding task. ; (44) Detect monthly delivery volume Is it less than If the detection result is yes, the corresponding task is identified as having low activity; if the detection result is no, proceed to step (45). (45) Calculate the monthly delivery volume and First difference and monthly delivery volume and The second difference is used to determine the activity level. If the first difference is less than the second difference, the corresponding task is considered to have high activity. If the first difference is greater than the second difference, the corresponding task is considered to have low activity.

6. The project task popularity evaluation method as described in claim 5, characterized in that, Poisson observation model Specifically as follows: ； in, , This represents the eigenvectors after orthogonal processing. The feature vector after intensity normalization. Representing the eigenvector With feature vectors The intensity of the inner product projection. These are adjustment parameters used to adjust the observation sensitivity. Let be the dimension of the feature vector corresponding to the task. For strength parameters, This indicates the monthly delivery volume for the corresponding task.

7. The project task popularity evaluation method as described in claim 5, characterized in that, Bernoulli observation model Specifically as follows: ； in, Indicates the dose attenuation factor. , Representing the eigenvector With feature vectors The intensity of the inner product projection. This indicates the monthly delivery volume for the corresponding task.

8. A project task popularity assessment device, characterized in that, The device includes: The task feature extraction module, orthogonal processing module, intensity standardization module, and activity prediction module are connected sequentially. The task feature extraction module is used to split the project task text into multi-task description text and extract the task keywords and feature vectors of each task. The orthogonal processing module is used to orthogonalize the feature vectors of the task. The intensity normalization module is used to normalize the intensity of the orthogonalized feature vectors. The activity prediction module is used to predict the activity level of the corresponding task based on the orthogonalized feature vectors and the intensity-normalized feature vectors.

9. The project task heat assessment device as described in claim 8, characterized in that, The activity prediction module includes: a Poisson observation model and a Bernoulli observation model connected to the orthogonal processing module and the intensity standardization module, as well as a prediction unit connected to the Poisson observation model and the Bernoulli observation model; The orthogonal processing module and the intensity normalization module respectively process the feature vectors of each task. eigenvectors Input a Poisson observation model, and the Poisson observation model will output the monthly delivery volume for the corresponding task. Predicting the monthly delivery volume of the unit. Is it greater than If the detection result is yes, the corresponding task is considered to have high activity; if the detection result is no, the orthogonal processing module and the intensity standardization module respectively process the feature vectors of each task. eigenvectors Input the Bernoulli observation model, and the Bernoulli observation model will output the monthly delivery volume for the corresponding task. Predicting monthly delivery volume of unit inspection Is it less than If the detection result is yes, the corresponding task is considered to have low activity; if the detection result is no, the monthly delivery volume is calculated. and First difference and monthly delivery volume and The second difference is used to determine the activity level. If the first difference is less than the second difference, the corresponding task is considered to have high activity. If the first difference is greater than the second difference, the corresponding task is considered to have low activity.

10. The project task popularity assessment device as described in claim 8, characterized in that, The orthogonal processing module performs orthogonalization of the task's feature vectors, and the specific processing procedure is as follows: Set a set of orthogonal bases according to the task requirements of each task. The feature vectors of the task are orthogonalized based on orthogonal bases, and the specific formula is as follows: ； in, Let represent the feature vector of the i-th task keyword. Let j-th task keyword be an orthogonal basis. For feature vectors The eigenvectors after orthogonal processing.

Images