Editing method and apparatus for description text

Abstract

The method comprises the following steps: obtaining a set of points based on historical information corresponding to a target object; extracting a plurality of sets of sentences from at least one explanation text corresponding to the target object based on the plurality of points in the set of points; determining a plurality of weight values corresponding to the plurality of points based on a user portrait corresponding to a target user, and determining the order of the plurality of points based on the weight values; sorting the sets of sentences corresponding to the plurality of points of the target object in the order to obtain a description text corresponding to the target user of the target object; and obtaining a long text through the extracted sets of sentences, and determining the order of the sets of sentences based on the weight values determined based on the user portrait, thereby enhancing the interaction with the user and realizing personalized generation of the corresponding description text for the target user.

Description

Technical Field

[0001] This disclosure relates to the field of text generation technology, and in particular to a method and apparatus for describing text editing. Background Technology

[0002] Text generation technology has been applied in various professional fields, such as dialogue systems and news script generation, becoming one of the important research tasks in the field of natural language processing. Text generation involves enabling machines to learn grammatical rules and expression habits from large amounts of natural language corpora, ultimately automatically generating text content that meets business objectives. However, intelligent applications increasingly emphasize the user experience of human-computer interaction, requiring more interactive and diverse text generation results to enhance interactivity and engagement with customers.

[0003] Currently, the most common text generation methods in the industry are template-based. Template-based methods are frequently used in the early stages of business operations, allowing for rapid deployment and are suitable for short text generation. However, they rely on manual annotation and require offline construction and maintenance of a large amount of rules and template knowledge, resulting in significant human resource costs. For applications in vertical domains, expert experience becomes particularly important, but this experience is often very scarce, and cultivating such experience requires a substantial time investment. Summary of the Invention

[0004] To address the aforementioned technical problems, this disclosure is proposed. Embodiments of this disclosure provide a method and apparatus for editing descriptive text.

[0005] According to one aspect of the present disclosure, a method for editing descriptive text is provided, comprising:

[0006] Based on the historical information corresponding to the target object, a set of lecture points including multiple lecture points is obtained;

[0007] Based on multiple lecture points in the lecture point set, multiple sentence sets are extracted from at least one explanatory text corresponding to the target object; wherein, each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence;

[0008] Based on the user profile corresponding to the target user, determine multiple weight values ​​corresponding to the multiple presentation points, and determine the order of the multiple presentation points based on the weight values;

[0009] According to the aforementioned order, the sentence sets corresponding to multiple points of the target object are sorted to obtain the description text of the target object corresponding to the target user.

[0010] Optionally, the step of obtaining a set of lecture points, including multiple lecture points, based on the basic information corresponding to the target object includes:

[0011] Entity recognition is performed on the historical information to determine multiple triples corresponding to the target object; wherein each triple includes an entity, an attribute corresponding to the entity, and an attribute value corresponding to the attribute;

[0012] Based on the association relationships between the multiple entities, the connection relationships between the multiple triples are determined;

[0013] Based on the multiple triples with connection relationships, the set of lecture points is determined; wherein each entity corresponds to one lecture point.

[0014] Optionally, before determining the connection relationship between the multiple triples based on the association relationship between the multiple entities, the method further includes:

[0015] Based on preset rules, the basic information corresponding to the target object is processed to obtain the rule attributes corresponding to at least one entity included in the target object;

[0016] Based on the rule attribute and the entity corresponding to the rule attribute, a triplet with a preset attribute value is formed, and at least one triplet is obtained.

[0017] Optionally, the step of extracting multiple sentence sets from at least one explanatory text corresponding to the target object based on multiple lecture points in the lecture point set includes:

[0018] Obtain at least one explanatory text corresponding to the target object;

[0019] Based on the teaching points corresponding to the teaching sentences included in the at least one explanatory text, the multiple teaching sentences are aggregated to obtain multiple sets corresponding to the multiple teaching points; wherein, each set includes at least one teaching sentence;

[0020] At least one of the explanatory sentences is extracted from each of the multiple sets to obtain the multiple sentence sets.

[0021] Optionally, the aggregation of the plurality of explanatory sentences based on the explanatory points corresponding to the explanatory sentences included in the at least one explanatory text includes:

[0022] Based on the multiple points of explanation, the explanatory sentences included in the explanatory text are classified and processed to determine the point of explanation corresponding to each explanatory sentence in the at least one explanatory sentence;

[0023] Based on the teaching point corresponding to each of the teaching sentences, determine the set corresponding to each teaching point.

[0024] Optionally, the step of extracting at least one of the explanatory sentences from the plurality of sets to obtain the plurality of sentence sets includes:

[0025] For each of the plurality of sets, each explanatory sentence in the set is determined to be a first expression sentence or a second expression sentence by matching, thereby obtaining a first expression set and a second expression set; wherein, the first expression set includes at least one first expression sentence, and the second expression set includes at least one second expression sentence;

[0026] Clustering is performed on the first expression set and the second expression set respectively to obtain multiple cluster sets;

[0027] One explanatory sentence is extracted from each of the plurality of cluster sets to obtain the plurality of sentence sets.

[0028] Optionally, the step of clustering the first expression set and the second expression set to obtain multiple cluster sets includes:

[0029] The core information of each first expression sentence is determined by dependency parsing. Based on the core information, the first expression set is clustered to obtain at least one cluster set. Each cluster set includes the explanatory sentences of at least one expression mode corresponding to the same core information.

[0030] Each of the second expression sentences is subjected to vector analysis using a preset network model to obtain at least one sentence vector. The second expression set is then clustered based on the at least one sentence vector to obtain at least one cluster set.

[0031] Optionally, determining multiple weight values ​​corresponding to the multiple presentation points based on the user profile corresponding to the target user, and determining the order of the multiple presentation points based on the weight values, includes:

[0032] Based on the order of the lecture points in the historical information, multiple weight values ​​are assigned to the multiple lecture points;

[0033] Obtain a user profile corresponding to the target user, and adjust at least one of the multiple weight values ​​based on the user profile;

[0034] The order of the multiple presentation points is determined based on the adjusted weight values.

[0035] Optionally, after determining the order of the plurality of lecture points based on the adjusted weight values, the method further includes:

[0036] In response to receiving an input instruction to adjust the order of presentation points, the order of the multiple presentation points is adjusted based on the instruction to obtain the adjusted order.

[0037] According to another aspect of the embodiments of this disclosure, an editing device for descriptive text is provided, comprising:

[0038] The lecture point acquisition module is used to obtain a lecture point set including multiple lecture points based on the historical information corresponding to the target object;

[0039] The sentence extraction module is used to extract multiple sentence sets from at least one explanatory text corresponding to the target object based on multiple lecture points in the lecture point set; wherein each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence;

[0040] The order determination module is used to determine multiple weight values ​​corresponding to the multiple presentation points based on the user profile corresponding to the target user, and to determine the order of the multiple presentation points based on the weight values;

[0041] The text determination module is used to sort the set of sentences corresponding to multiple points of the target object in the order stated above, so as to obtain the description text of the target object corresponding to the target user.

[0042] Optionally, the lecture point acquisition module includes:

[0043] The triplet unit is used to perform entity recognition on the historical information and determine multiple triplets corresponding to the target object; wherein each triplet includes an entity, an attribute corresponding to the entity, and an attribute value corresponding to the attribute;

[0044] A group association unit is used to determine the connection relationship between the multiple triples based on the association relationship between the multiple entities;

[0045] The lecture point set unit is used to determine the lecture point set based on the multiple triples with connection relationships; wherein each entity corresponds to one lecture point.

[0046] Optionally, the lecture point acquisition module further includes:

[0047] The rule attribute unit is used to process the basic information corresponding to the target object based on preset rules to obtain the rule attributes corresponding to at least one entity included in the target object.

[0048] An attribute group unit is used to form a triple with a preset attribute value based on the rule attribute and the entity corresponding to the rule attribute, thereby obtaining at least one triple.

[0049] Optionally, the sentence extraction module includes:

[0050] The text acquisition unit is used to acquire at least one explanatory text corresponding to the target object;

[0051] An aggregation unit is used to aggregate the plurality of explanatory sentences based on the explanatory points corresponding to the explanatory sentences included in the at least one explanatory text, to obtain a plurality of sets corresponding to the plurality of explanatory points; wherein each set includes at least one explanatory sentence;

[0052] The sentence extraction unit is used to extract at least one of the explanatory sentences from the plurality of sets respectively to obtain the plurality of sentence sets.

[0053] Optionally, the aggregation unit is specifically used to classify the explanatory sentences included in the explanatory text based on the plurality of explanatory points, determine the explanatory point corresponding to each of the at least one explanatory sentences, and determine the set corresponding to each explanatory point based on the explanatory point corresponding to each explanatory sentence.

[0054] Optionally, the sentence extraction unit is specifically used to determine, for each of the plurality of sets, whether each explanatory sentence in the set is a first expression sentence or a second expression sentence by matching, thereby obtaining a first expression set and a second expression set; wherein, the first expression set includes at least one first expression sentence, and the second expression set includes at least one second expression sentence; the first expression set and the second expression set are clustered respectively to obtain a plurality of cluster sets; and an explanatory sentence is extracted from each of the plurality of cluster sets to obtain the plurality of sentence sets.

[0055] Optionally, when the sentence extraction unit clusters the first expression set and the second expression set to obtain multiple cluster sets, it is used to determine the core information of each first expression sentence through dependency parsing, and based on the core information, cluster the first expression set to obtain at least one cluster set; wherein, each cluster set includes the explanatory sentences of at least one expression mode corresponding to the same core information; and to perform vector analysis on each second expression sentence through a preset network model to obtain at least one sentence vector, and to cluster the second expression set based on the at least one sentence vector to obtain at least one cluster set.

[0056] Optionally, the order determination module is specifically used to assign multiple weight values ​​to the multiple lecture points based on the order of the lecture points in the historical information; obtain a user profile corresponding to the target user; adjust at least one of the multiple weight values ​​based on the user profile; and determine the order of the multiple lecture points based on the adjusted weight values.

[0057] Optionally, the sequence determination module is further configured to, in response to receiving an input instruction to adjust the order of the lecture points, adjust the order of the multiple lecture points based on the instruction to obtain the adjusted order.

[0058] According to another aspect of the present disclosure, an electronic device is provided, comprising:

[0059] Memory, used to store computer program products;

[0060] A processor is configured to execute a computer program product stored in the memory, and when the computer program product is executed, to implement the method for editing descriptive text as described in any of the above embodiments.

[0061] According to another aspect of the present disclosure, a computer-readable storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement the method for editing descriptive text as described in any of the above embodiments.

[0062] According to another aspect of the present disclosure, a computer program product is provided, including computer program instructions that, when executed by a processor, implement the method for editing descriptive text as described in any of the above embodiments.

[0063] Based on the above embodiments of this disclosure, a method and apparatus for editing descriptive text are provided. A set of lecture points, including multiple lecture points, is obtained based on historical information corresponding to a target object. Multiple sentence sets are extracted from at least one explanatory text corresponding to the target object based on the multiple lecture points in the lecture point set. Each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence. Multiple weight values ​​corresponding to the multiple lecture points are determined based on the user profile corresponding to the target user, and the order of the multiple lecture points is determined based on the weight values. The sentence sets corresponding to the multiple lecture points of the target object are sorted according to the order to obtain the descriptive text corresponding to the target user. By extracting sentence sets to obtain long text and determining the order of sentence sets based on the weight values ​​determined by the user profile, interaction with the user is enhanced, and personalized descriptive text is generated for the target user.

[0064] The technical solutions of this disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0065] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0066] Figure 1 This is a flowchart illustrating a method for editing descriptive text provided in an exemplary embodiment of this disclosure;

[0067] Figure 2 This is a public announcement Figure 1 A flowchart illustrating step 102 in the illustrated embodiment;

[0068] Figure 3 This is a public announcement Figure 1 A flowchart illustrating step 104 in the illustrated embodiment;

[0069] Figure 4 This is a public announcement Figure 3 A flowchart illustrating step 1043 in the illustrated embodiment;

[0070] Figure 5 This is a public announcement Figure 1 A flowchart illustrating step 106 in the illustrated embodiment;

[0071] Figure 6 This is a schematic diagram of the structure of a text editing device provided in an exemplary embodiment of the present disclosure;

[0072] Figure 7 This is a structural diagram of an electronic device provided in an exemplary embodiment of this disclosure. Detailed Implementation

[0073] Hereinafter, exemplary embodiments according to the present disclosure will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present disclosure, and not all embodiments of the present disclosure, and it should be understood that the present disclosure is not limited to the exemplary embodiments described herein.

[0074] It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of this disclosure.

[0075] Those skilled in the art will understand that the terms "first," "second," etc., in the embodiments of this disclosure are only used to distinguish different steps, devices, or modules, and do not represent any specific technical meaning, nor do they indicate a necessary logical order between them.

[0076] It should also be understood that in the embodiments disclosed herein, "a plurality of" may refer to two or more, and "at least one" may refer to one, two or more.

[0077] It should also be understood that any component, data or structure mentioned in the embodiments of this disclosure can generally be understood as one or more unless expressly defined or given to the contrary in the context.

[0078] Furthermore, the term "and / or" in this disclosure is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this disclosure generally indicates that the preceding and following related objects have an "or" relationship. The data referred to in this disclosure can include unstructured data such as text, images, and videos, as well as structured data.

[0079] It should also be understood that the description of the various embodiments in this disclosure emphasizes the differences between the various embodiments, and the similarities or similarities can be referred to each other. For the sake of brevity, they will not be described in detail.

[0080] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0081] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this disclosure or its application or use.

[0082] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0083] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0084] The embodiments disclosed herein can be applied to electronic devices such as terminal devices, computer systems, and servers, and can operate together with a wide range of other general-purpose or special-purpose computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments, and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, and servers include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputer systems, mainframe computer systems, and distributed cloud computing environments including any of the above systems, etc.

[0085] Electronic devices such as terminal devices, computer systems, and servers can be described in the general context of computer system executable instructions (such as program modules) executed by a computer system. Typically, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in distributed cloud computing environments, where tasks are executed by remote processing devices linked through communication networks. In distributed cloud computing environments, program modules can reside on local or remote computing system storage media, including storage devices.

[0086] Exemplary methods

[0087] Figure 1 This is a flowchart illustrating a method for editing descriptive text provided in an exemplary embodiment of this disclosure. This embodiment can be applied to electronic devices, such as… Figure 1 As shown, it includes the following steps:

[0088] Step 102: Based on the historical information corresponding to the target object, obtain a set of lecture points including multiple lecture points.

[0089] In this embodiment, the target object can be a describable product such as a property or a vehicle; historical information can be existing information describing the target object obtained prior to this embodiment. For example, when the target object is a property, historical information can be the agent's viewing data or the agent's online communication description of the property, etc. That is, historical information is existing information describing the target object, and the acquisition methods can include, but are not limited to: obtaining from a database of pre-stored historical information and / or web crawling, etc.; the description point can be a describable direction of the target object. For example, when the target object is a property, the description point can include location, floor, surrounding environment, orientation, etc.; this embodiment mines all describable directions of the target object based on historical information.

[0090] Step 104: Based on multiple points in the lecture point set, extract multiple sentence sets from at least one explanatory text corresponding to the target object.

[0091] Each sentence set corresponds to a teaching point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence.

[0092] In one embodiment, the explanatory text is the basis for the current editing of the descriptive text. It can be the obtained text that describes the target object by humans, or other audio, video or text materials that need to be organized. When the gesture is in the form of audio or video, the explanatory text corresponding to the audio or video material can be obtained through technologies such as speech recognition. By extraction, a set of sentences describing each of the multiple points can be obtained from at least one explanatory text.

[0093] Step 106: Determine multiple weight values ​​for multiple presentation points based on the user profile corresponding to the target user, and determine the order of multiple presentation points based on the weight values.

[0094] In this embodiment, the characteristics of the target user are described through user profiles, which distinguishes the target user from other users. The lecture points are sorted based on the weight values ​​determined by the user profiles, so that the sorted lecture point order better meets the needs of the target user, thus realizing personalized explanation.

[0095] Step 108: Sort the sentence sets corresponding to multiple points of the target object in order to obtain the description text of the target user corresponding to the target object.

[0096] This embodiment connects the sentence sets corresponding to multiple lecture points in the order determined by the user profile, so that the sentence sets corresponding to multiple lecture points form a long text description text in sequence, which solves the problems of not being able to obtain long text and the text not being strongly related to the user in the prior art.

[0097] This disclosure provides a method for editing descriptive text, which, based on historical information corresponding to a target object, obtains a set of lecture points including multiple lecture points; based on the multiple lecture points in the lecture point set, extracts multiple sentence sets from at least one explanatory text corresponding to the target object; wherein each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence; based on the user profile corresponding to the target user, determines multiple weight values ​​corresponding to the multiple lecture points, and determines the order of the multiple lecture points based on the weight values; sorts the sentence sets corresponding to the multiple lecture points of the target object according to the order, thereby obtaining the descriptive text corresponding to the target user; by extracting sentence sets to obtain long text, and by determining the order of sentence sets based on the weight values ​​determined by the user profile, the interaction with the user is enhanced, and personalized descriptive text is generated for the target user.

[0098] like Figure 2 As shown above, in the above Figure 1 Based on the illustrated embodiment, step 102 may include the following steps:

[0099] Step 1021: Perform entity recognition on historical information to determine multiple triples corresponding to the target object.

[0100] Each triple consists of an entity, the entity's corresponding attribute, and the attribute's corresponding value.

[0101] In this embodiment, a triple is formed by an entity, the attribute corresponding to the entity, and the attribute value corresponding to the attribute. For example, when the target object is a house, the entity can be "living room", the attribute can be "area", and the attribute value is the specific area value of the living room, such as "18 square meters". In addition, each entity can correspond to at least one attribute, and multiple different entities may have the same attribute. For example, "bedroom", "living room" and "bathroom" all have area attributes, but the corresponding attribute values ​​are different.

[0102] Step 1022: Based on the association relationships between multiple entities, determine the connection relationships between multiple triples.

[0103] Since multiple entities are all parts of the target object, there must be relationships between them. For example, if an entity is a component of the target object, then the relationship between multiple entities is the connection structure between the multiple components. When the target object is a house, the entities are parts of the house structure. In this case, the relationship between entities can be obtained by analyzing the floor plan of the house, or by determining the relationship by the order of events in the viewing materials. For example, if the floor plan shows that the living room has a balcony and is connected to the balcony, then the triple with the living room as the entity can be determined to have a connection relationship with the triple with the balcony as the entity.

[0104] Step 1023: Determine the set of lecture points based on multiple triples with connection relationships.

[0105] Each entity corresponds to one point in the presentation.

[0106] In this embodiment, multiple triples corresponding to the target object are connected through connection relationships to form a knowledge graph-like set of lecture points. This set of lecture points includes all lecture points that can describe the target object, and these lecture points have a set of association relationships. In this embodiment, the relationship between knowledge points can be understood more intuitively through the form of a knowledge graph.

[0107] Optionally, based on the above embodiments, before step 1022, the following may also be included:

[0108] Based on preset rules, the basic information corresponding to the target object is processed to obtain the rule attributes corresponding to at least one entity included in the target object;

[0109] Based on the rule attributes and the entities corresponding to the rule attributes, triples with preset attribute values ​​are formed, resulting in at least one triplet.

[0110] In this embodiment, the basic information can be the original information of the target object. For example, when the target object is a house, the basic information can be a floor plan; when the target object is a vehicle, the basic information can be design drawings, etc. Through preset inference rules, the corresponding rule attributes are determined. For example, based on a certain structural information in the basic information, the corresponding rule attribute is determined through inference rules. For instance, based on the structure in the floor plan where windows are on both the south and north sides, the house is determined to have the rule attribute of "north-south ventilation" according to the inference rules. Based on the floor information given in the floor plan and the total number of floors in the corresponding building, the house is determined to have the rule attribute of "north-south ventilation". The rule attribute is either "high floor" or "low floor". Usually, the rule attribute corresponds to only a limited number of attribute values. Therefore, in this embodiment, having the rule attribute is used as the preset attribute value, that is, the preset attribute value is "has" or "is". For example, a house has the rule attribute of being north-south facing. The corresponding entity can be the target object. The target object as a whole is described as an entity. For example, the resulting triple is: Property A - is - north-south facing. By adding the triple corresponding to the rule attribute to step 1022, the connection relationship between the subject corresponding to the rule attribute and other subjects is established, thereby forming a knowledge graph-style set of lecture points.

[0111] like Figure 3 As shown above, in the above Figure 1 Based on the illustrated embodiment, step 104 may include the following steps:

[0112] Step 1041: Obtain at least one explanatory text corresponding to the target object.

[0113] Optionally, at least one explanatory text can be obtained by processing audio, video, or text materials describing the target object. For audio or video materials, text extraction can be achieved through technologies such as speech recognition (deep neural networks, etc.). At least one explanatory text can be obtained based on the obtained text. For example, when the target object is a house, text extraction can be performed based on the audio or video of the guided tour to obtain at least one corresponding explanatory text. Optionally, one audio of the guided tour can correspond to one explanatory text.

[0114] Step 1042: Based on the teaching points corresponding to the teaching sentences included in at least one teaching text, aggregate multiple teaching sentences to obtain multiple sets corresponding to multiple teaching points.

[0115] Each set includes at least one explanatory sentence.

[0116] Optionally, the explanatory sentences included in the explanatory text are classified based on multiple explanatory points to determine the explanatory point corresponding to each explanatory sentence in at least one explanatory sentence; based on the explanatory points corresponding to each explanatory sentence, a set corresponding to each explanatory point is determined.

[0117] This embodiment achieves aggregation of multiple explanatory sentences by determining the teaching point corresponding to each explanatory sentence. All explanatory sentences in the aggregated set correspond to the same teaching point. By aggregating the explanatory sentences corresponding to each teaching point into a set, it is easier to identify repetitive explanatory content and avoid the problem of some teaching points being explained repeatedly while others are ignored in the final generated description text.

[0118] Step 1043: Extract at least one explanatory sentence from each of the multiple sets to obtain multiple sentence sets.

[0119] This embodiment extracts explanatory sentences from the sets corresponding to each lecture point, ensuring that the obtained sentence set includes explanatory sentences for each lecture point across multiple lecture points. This avoids the problem of some lecture points being missed due to random extraction, and ensures that each lecture point is reflected in the description text. This allows the description text to provide a more comprehensive description of the target object, enabling the target user to gain a complete understanding of the target object through the description text.

[0120] like Figure 4 As shown above, in the above Figure 3 Based on the illustrated embodiment, for each of the plurality of sets, step 1043 may include the following steps:

[0121] Step 401: By matching, determine whether each explanatory sentence in the set is the first expression sentence or the second expression sentence, and obtain the first expression set and the second expression set.

[0122] The first set of expressions includes at least one first expression sentence, and the second set of expressions includes at least one second expression sentence.

[0123] Optionally, the first expression can be a factual statement, and the second expression can be a personalized description. The factual statement refers to the triplet information processed and reasoned from the set, used to explain the basic content of the house. This is known and definite information, and the sentence structure is relatively simple. The personalized description is a further explanation of the factual statement; the style varies from person to person, exhibiting diversity (e.g., described using different descriptive languages) and a high degree of freedom. This embodiment, for each point of discussion, achieves diverse descriptions of the points of discussion by mining both factual statements and personalized descriptions.

[0124] Step 402: Cluster the first expression set and the second expression set respectively to obtain multiple cluster sets.

[0125] In this embodiment, to avoid repeated expression of the same content, the first expression set and the second expression set are clustered based on the specific content. Each cluster set includes at least one first expression sentence or at least one second expression sentence, and all first expression sentences or all second expression sentences in each cluster set correspond to the same specific content.

[0126] Step 403: Extract one explanatory sentence from each of the multiple cluster sets to obtain multiple sentence sets.

[0127] This embodiment extracts one explanatory sentence from each cluster set; that is, it obtains a first explanatory sentence and a second explanatory sentence for each specific content. Each specific content is described using only one first explanatory sentence and one second explanatory sentence, thus avoiding the problem of repeated content. The extraction is random. Each point of explanation can correspond to at least one first explanatory sentence and at least one second explanatory sentence. Through all the obtained explanatory sentences, a sentence set corresponding to multiple points of explanation can be obtained. This sentence set includes all the descriptive content that completely describes the target object. The explanatory text generated from this sentence set is the long text corresponding to the target object, overcoming the problem of not being able to generate long text in the prior art.

[0128] Optionally, step 402 may include:

[0129] The core information of each first expression sentence is determined by dependency parsing. Based on the core information, the first expression set is clustered to obtain at least one cluster set. Each cluster set includes explanatory sentences of at least one expression mode corresponding to the same core information.

[0130] By performing vector analysis on each second expression sentence using a pre-defined network model, at least one sentence vector is obtained. Based on at least one sentence vector, the second expression set is clustered to obtain at least one cluster set.

[0131] In this embodiment, dependency parsing is used to extract the core information (e.g., only including subject, predicate, object, or the main content of the sentence) from the first expression sentence. Aggregation based on this core information yields multiple expressions for the same core information (e.g., different adverbs, different word order, etc.). Aggregation based on core information ignores the impact of different knowledge enumeration values ​​on the expression. In this embodiment, the mining of the second expression sentence mainly involves aggregating multiple second expression sentences using sentence vectors. Specifically, a pre-set network model (e.g., a trained model for vector extraction of the second expression sentences, expressing the second expression sentences based on the vectors) can be used to implement this function using any existing technology. The deep neural network model is used (this embodiment does not limit the structure of the network model used) to vectorize the second expression sentence, obtaining the sentence vector corresponding to each second expression sentence. Based on the distance between multiple sentence vectors (the closer the distance, the more similar the content of the corresponding second expression sentence), multiple sentence vectors are clustered (sentence vectors with a distance less than a set value are determined to be in the same cluster set, the set value can be set according to the actual application scenario). The resulting at least one cluster set (the content similarity between multiple second expression sentences in each cluster set reaches the set condition) includes diverse expressions of the same content in each cluster set, avoiding the repetition of similar content in the descriptive text.

[0132] like Figure 5 As shown above, in the above Figure 1 Based on the illustrated embodiment, step 106 may include the following steps:

[0133] Step 1061: Assign multiple weight values ​​to multiple lecture points based on the order of the lecture points in the historical information.

[0134] In this embodiment, the target objects are initially sorted in the historical information. According to this sorting, different weight values ​​can be assigned to multiple lecture points (for example, the order between multiple lecture points can be represented by a probability transition matrix. For example, in 10 lectures, the second lecture point is explained after the first lecture point 4 times, and the third lecture point is explained after the first lecture point 6 times. Then the probability of transferring from the first lecture point to the second lecture point is 0.4, and the probability of transferring from the first lecture point to the third lecture point is 0.6. When assigning weight values, the first lecture point has the largest weight value, the third lecture point has the second largest weight value, and the second lecture point has the smallest weight value. The lecture point order can be determined by the path with the largest sum of probabilities obtained by connecting all lecture points in the probability transition graph; or the largest probability value can be selected each time a lecture point is transferred to another lecture point, and the final path that traverses all lecture points can be used as the lecture point order. Optionally, higher weight values ​​can be assigned to lecture points that are ranked earlier, and lower weight values ​​can be assigned to lecture points that are ranked later, or vice versa.

[0135] Step 1062: Obtain the user profile corresponding to the target user, and adjust at least one of the multiple weight values ​​based on the user profile.

[0136] Optionally, user profiles may include, but are not limited to, basic attribute data, preference data, and behavioral data. Attribute data can be obtained based on information registered or input by the target user; preference data can be obtained based on information browsed by the target user; and behavioral data can be obtained based on the target user's actions on objects. For example, preference data can be obtained by determining that the target user prefers properties with elevators through browsing multiple properties. In this case, the weight value of the elevator information corresponding to the target object can be adjusted (increasing the weight value when sorting from largest to smallest, and decreasing the weight value when sorting from smallest to largest, specifically by controlling the order of the information to move forward), so that the target user can browse elevator information first, increasing the user's interest in the property.

[0137] Step 1063: Determine the order of multiple presentation points based on the adjusted weight values.

[0138] In this embodiment, the weight value of at least one of the multiple presentation points is adjusted based on user profiling, prioritizing presentation points that users are more interested in. This increases user engagement with the descriptive text and reduces the likelihood of users not reading the entire description due to weak interest. Furthermore, personalized descriptive texts are provided to different users through user profiling, avoiding monotonous descriptions of the target audience and increasing the presenter's interest, preventing decreased motivation from repetitive explanations. Additionally, user preferences are collected during the presentation process. For example, user-input commands to adjust the order of presentation points are used to collect the forward-ordered presentation points as user preferences, enabling dynamic updates to presentation point weights and real-time generation of diverse presentation texts.

[0139] Optionally, based on the above embodiment, after performing step 1063, the following may be included:

[0140] In response to receiving an input instruction to adjust the order of presentation points, the order of multiple presentation points is adjusted based on the instruction to obtain the adjusted order.

[0141] In this embodiment, the instruction to adjust the order of presentation points can be issued by the target user or the presenter (e.g., a real estate agent explaining a property). By receiving the instruction to adjust the order of presentation points, the needs of the target user or presenter can be received in real time, and the user's latest preferences can be obtained (input instructions are a type of behavioral data). By receiving user instructions, the user profile can be enriched, and the order of presentation points can be adjusted in a targeted manner so that the adjusted order is more in line with the preferences of the target user or presenter. In addition, when there is a conflict between the instructions of the target user and the presenter, the instructions of the target user shall prevail.

[0142] Any of the descriptive text editing methods provided in this disclosure can be executed by any suitable device with data processing capabilities, including but not limited to: terminal devices and servers. Alternatively, any of the descriptive text editing methods provided in this disclosure can be executed by a processor, such as by a processor executing any of the descriptive text editing methods mentioned in this disclosure by calling corresponding instructions stored in memory. Further details will not be elaborated below.

[0143] Exemplary device

[0144] Figure 6 This is a schematic diagram of the structure of a text editing device provided in an exemplary embodiment of this disclosure. Figure 6 As shown, the apparatus provided in this embodiment includes:

[0145] The lecture point acquisition module 61 is used to obtain a lecture point set including multiple lecture points based on the historical information corresponding to the target object.

[0146] The sentence extraction module 62 is used to extract multiple sentence sets from at least one explanatory text corresponding to the target object based on multiple explanatory points in the explanatory point set.

[0147] Each sentence set corresponds to a teaching point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence.

[0148] The sequence determination module 63 is used to determine multiple weight values ​​corresponding to multiple lecture points based on the user profile corresponding to the target user, and to determine the order of multiple lecture points based on the weight values.

[0149] The text determination module 64 is used to sort the sentence sets corresponding to multiple points of the target object in order to obtain the description text of the target user corresponding to the target object.

[0150] This disclosure provides a descriptive text editing device according to the above embodiments. Based on historical information corresponding to a target object, a set of lecture points including multiple lecture points is obtained. Based on the multiple lecture points in the lecture point set, multiple sentence sets are extracted from at least one explanatory text corresponding to the target object. Each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence. Multiple weight values ​​corresponding to the multiple lecture points are determined based on the user profile corresponding to the target user, and the order of the multiple lecture points is determined based on the weight values. The sentence sets corresponding to the multiple lecture points of the target object are sorted according to the order to obtain the descriptive text corresponding to the target user. By extracting sentence sets to obtain long text and determining the order of sentence sets based on the weight values ​​determined by the user profile, the interaction with the user is enhanced, and personalized descriptive text is generated for the target user.

[0151] Optionally, the lecture point acquisition module 61 includes:

[0152] The triplet unit is used to identify entities from historical information and determine multiple triples corresponding to the target object; each triplet includes an entity, the entity's corresponding attribute, and the attribute value corresponding to the attribute.

[0153] Group association unit, used to determine the connection relationship between multiple triples based on the association relationship between multiple entities;

[0154] The lecture point set unit is used to determine the lecture point set based on multiple triples with connection relationships; where each entity corresponds to one lecture point.

[0155] Optionally, the lecture point acquisition module 61 also includes:

[0156] The rule attribute unit is used to process the basic information corresponding to the target object based on preset rules to obtain the rule attributes corresponding to at least one entity included in the target object.

[0157] The attribute group unit is used to form a triple with a preset attribute value based on the rule attribute and the entity corresponding to the rule attribute, so as to obtain at least one triple.

[0158] Optionally, the sentence extraction module 62 includes:

[0159] The text acquisition unit is used to obtain at least one explanatory text corresponding to the target object;

[0160] An aggregation unit is used to aggregate multiple explanatory sentences based on the explanatory points corresponding to the explanatory sentences included in at least one explanatory text, to obtain multiple sets corresponding to multiple explanatory points; wherein each set includes at least one explanatory sentence;

[0161] The sentence extraction unit is used to extract at least one explanatory sentence from multiple sets to obtain multiple sentence sets.

[0162] Optionally, the aggregation unit is specifically used to classify the explanatory sentences included in the explanatory text based on multiple explanatory points, determine the explanatory points corresponding to each explanatory sentence in at least one explanatory sentence, and determine the set corresponding to each explanatory point based on the explanatory points corresponding to each explanatory sentence.

[0163] Optionally, the sentence extraction unit is specifically used to determine, for each of the multiple sets, whether each explanatory sentence in the set is a first expression sentence or a second expression sentence by matching, thereby obtaining a first expression set and a second expression set; wherein, the first expression set includes at least one first expression sentence, and the second expression set includes at least one second expression sentence; the first expression set and the second expression set are clustered respectively to obtain multiple cluster sets; and an explanatory sentence is extracted from each of the multiple cluster sets to obtain multiple sentence sets.

[0164] Optionally, when the sentence extraction unit clusters the first expression set and the second expression set to obtain multiple cluster sets, it is used to determine the core information of each first expression sentence through dependency parsing, and based on the core information, to cluster the first expression set to obtain at least one cluster set; wherein, each cluster set includes explanatory sentences of at least one expression mode corresponding to the same core information; and to perform vector analysis on each second expression sentence through a preset network model to obtain at least one sentence vector, and to cluster the second expression set based on at least one sentence vector to obtain at least one cluster set.

[0165] Optionally, the sequence determination module 63 is specifically used to assign multiple weight values ​​to multiple lecture points based on the order of lecture points in historical information; obtain the user profile corresponding to the target user; adjust at least one of the multiple weight values ​​based on the user profile; and determine the order of multiple lecture points based on the adjusted weight values.

[0166] Optionally, the sequence determination module 63 is further configured to, in response to receiving an input instruction to adjust the order of multiple lecture points, adjust the order of the lecture points based on the instruction to obtain the adjusted order.

[0167] Exemplary electronic devices

[0168] Below, for reference Figure 7 This describes an electronic device according to embodiments of the present disclosure. The electronic device may be either or both of a first device 100 and a second device 200, or a standalone device independent of them, which may communicate with the first and second devices to receive acquired input signals from them.

[0169] Figure 7 A block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0170] like Figure 7 As shown, the electronic device 70 includes one or more processors 71 and a memory 72.

[0171] The processor 71 may be a central processing unit (CPU) or other form of processing unit with data processing and / or instruction execution capabilities, and may control other components in the electronic device 70 to perform desired functions.

[0172] The memory can store one or more computer program products, and the memory can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program products can be stored on the computer-readable storage medium, and the processor can run the computer program products to implement the methods for editing the descriptive text of the various embodiments of this disclosure described above, and / or other desired functions.

[0173] In one example, the electronic device 70 may also include an input device 73 and an output device 74, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0174] For example, when the electronic device is a first device 100 or a second device 200, the input device 73 can be the aforementioned microphone or microphone array for capturing the input signal from the sound source. When the electronic device is a standalone device, the input device 73 can be a communication network connector for receiving the acquired input signals from the first device 100 and the second device 200.

[0175] In addition, the input device 73 may also include, for example, a keyboard, a mouse, etc.

[0176] The output device 74 can output various information to the outside, including determined distance information, direction information, etc. The output device 74 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0177] Of course, for the sake of simplicity, Figure 7 Only some of the components of the electronic device 70 relevant to this disclosure are shown, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device 70 may include any other suitable components depending on the specific application.

[0178] Exemplary computer program products and computer-readable storage media

[0179] In addition to the methods and apparatus described above, embodiments of this disclosure may also be computer program products comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods for editing descriptive text according to various embodiments of this disclosure as described in the foregoing portions of this specification.

[0180] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0181] Furthermore, embodiments of this disclosure may also be computer-readable storage media storing computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods for editing descriptive text according to various embodiments of this disclosure as described in the "Exemplary Methods" section above.

[0182] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0183] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0184] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For system embodiments, since they largely correspond to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0185] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0186] The methods and apparatus of this disclosure may be implemented in many ways. For example, they may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order of steps for the methods is for illustrative purposes only, and the steps of the methods of this disclosure are not limited to the order specifically described above unless otherwise specifically stated. Furthermore, in some embodiments, this disclosure may also be implemented as a program recorded on a recording medium, the program including machine-readable instructions for implementing the methods according to this disclosure. Thus, this disclosure also covers recording media storing programs for performing the methods according to this disclosure.

[0187] It should also be noted that in the apparatus, devices, and methods of this disclosure, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions to this disclosure.

[0188] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0189] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A method for editing text, characterized in that, include: Based on the historical information corresponding to the target object, a set of lecture points including multiple lecture points is obtained; Based on multiple lecture points in the lecture point set, multiple sentence sets are extracted from at least one explanatory text corresponding to the target object; wherein, each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence; Based on the user profile corresponding to the target user, multiple weight values ​​are determined for the multiple presentation points, and the order of the multiple presentation points is determined based on the weight values; each presentation point corresponds to one weight value. According to the order described above, the sentence sets corresponding to multiple points of the target object are sorted to obtain the description text of the target object corresponding to the target user; The process involves extracting multiple sentence sets from at least one explanatory text corresponding to the target object based on multiple lecture points in the lecture point set, including: Obtain at least one explanatory text corresponding to the target object; Based on the teaching points corresponding to the teaching sentences included in the at least one explanatory text, the multiple teaching sentences are aggregated to obtain multiple sets corresponding to the multiple teaching points; wherein, each set includes at least one teaching sentence; For each of the plurality of sets, each explanatory sentence in the set is determined to be a first expression sentence or a second expression sentence by matching, thereby obtaining a first expression set and a second expression set; wherein, the first expression set includes at least one first expression sentence, and the second expression set includes at least one second expression sentence; Clustering is performed on the first expression set and the second expression set to obtain multiple cluster sets; each cluster set includes at least one of the first expression sentences or at least one of the second expression sentences. One explanatory sentence is extracted from each of the plurality of cluster sets to obtain the plurality of sentence sets.

2. The method according to claim 1, characterized in that, Based on the basic information corresponding to the target object, a set of lecture points is obtained, including multiple lecture points: Entity recognition is performed on the historical information to determine multiple triples corresponding to the target object; wherein each triple includes an entity, an attribute corresponding to the entity, and an attribute value corresponding to the attribute; Based on the association relationships between multiple entities, determine the connection relationships between the multiple triples; The set of lecture points is determined based on multiple triples with connection relationships; wherein each entity corresponds to one lecture point.

3. The method according to claim 2, characterized in that, Before determining the connection relationships between the multiple triples based on the association relationships between the multiple entities, the method further includes: Based on preset rules, the basic information corresponding to the target object is processed to obtain the rule attributes corresponding to at least one entity included in the target object; Based on the rule attribute and the entity corresponding to the rule attribute, a triplet with a preset attribute value is formed, and at least one triplet is obtained.

4. The method according to any one of claims 1-3, characterized in that, The aggregation of the plurality of explanatory sentences based on the explanatory points corresponding to the explanatory sentences included in the at least one explanatory text includes: Based on the multiple points of explanation, the explanatory sentences included in the explanatory text are classified and processed to determine the point of explanation corresponding to each explanatory sentence in the at least one explanatory sentence; Based on the teaching point corresponding to each of the teaching sentences, determine the set corresponding to each teaching point.

5. The method according to any one of claims 1-3, characterized in that, The first expression set and the second expression set are clustered respectively to obtain multiple cluster sets, including: The core information of each first expression sentence is determined by dependency parsing. Based on the core information, the first expression set is clustered to obtain at least one cluster set. Each cluster set includes the explanatory sentences of at least one expression mode corresponding to the same core information. Each of the second expression sentences is subjected to vector analysis using a preset network model to obtain at least one sentence vector. The second expression set is then clustered based on the at least one sentence vector to obtain at least one cluster set.

6. The method according to any one of claims 1-3, characterized in that, The step of determining multiple weight values ​​corresponding to the multiple presentation points based on the user profile corresponding to the target user, and determining the order of the multiple presentation points based on the weight values, includes: Based on the order of the lecture points in the historical information, multiple weight values ​​are assigned to the multiple lecture points; Obtain a user profile corresponding to the target user, and adjust at least one of the multiple weight values ​​based on the user profile; The order of the multiple presentation points is determined based on the adjusted weight values.

7. The method according to claim 6, characterized in that, After determining the order of the plurality of lecture points based on the adjusted weight values, the method further includes: In response to receiving an input instruction to adjust the order of presentation points, the order of the multiple presentation points is adjusted based on the instruction to obtain the adjusted order.

8. A text editing device, characterized in that, include: The lecture point acquisition module is used to obtain a lecture point set including multiple lecture points based on the historical information corresponding to the target object; The sentence extraction module is used to extract multiple sentence sets from at least one explanatory text corresponding to the target object based on multiple lecture points in the lecture point set; wherein each sentence set corresponds to one lecture point, each explanatory text includes multiple explanatory sentences, and each sentence set includes at least one explanatory sentence; The sequence determination module is used to determine multiple weight values ​​corresponding to the multiple lecture points based on the user profile corresponding to the target user, and to determine the order of the multiple lecture points based on the weight values; each lecture point corresponds to one weight value; The text determination module is used to sort the set of sentences corresponding to multiple points of the target object in the order described above, so as to obtain the description text of the target object corresponding to the target user. The sentence extraction module includes: The text acquisition unit is used to acquire at least one explanatory text corresponding to the target object; An aggregation unit is used to aggregate the plurality of explanatory sentences based on the explanatory points corresponding to the explanatory sentences included in the at least one explanatory text, to obtain a plurality of sets corresponding to the plurality of explanatory points; wherein each set includes at least one explanatory sentence; A sentence extraction unit is configured to, for each of the plurality of sets, determine, through matching, whether each explanatory sentence in the set is a first expression sentence or a second expression sentence, thereby obtaining a first expression set and a second expression set; wherein, the first expression set includes at least one first expression sentence, and the second expression set includes at least one second expression sentence; clustering is performed on the first expression set and the second expression set respectively to obtain a plurality of cluster sets; each cluster set includes at least one first expression sentence or includes at least one second expression sentence; and an explanatory sentence is extracted from each of the plurality of cluster sets to obtain the plurality of sentence sets.

Images