Smart phone outbound method and device for target post recruitment, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a smart phone outbound method and device for target post recruitment, electronic equipment and a storage medium. The method comprises the following steps: determining target question content associated with target terminal equipment; detecting whether standard reply content matched with the target question content exists in a preset knowledge storage library or not; if the standard reply content matched with the target question content exists, determining the standard reply content matched with the target question content as a target question answering result of the target question content; if the standard reply content matched with the target question content does not exist, searching a target corpus related to the target question content from a preset corpus; generating target cue word content based on the target corpus and the target question content, and outputting a target question-answering result of the target question content through a preset question-answering model; and performing voice conversion based on the target question-answer result and returning to the target terminal device. According to the technical scheme, the reply efficiency and accuracy during target post recruitment in a smart phone outbound call scene can be improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of Internet technology, and in particular to an intelligent telephone outbound calling method, device, electronic device and storage medium for recruiting target positions. Background Art

[0002] In today’s recruitment and many interactive scenarios, there are significant differences in different interaction modes. Most recruitment platforms mainly use text interaction to obtain or push information through text messaging, and the text interaction method has a loose time limit for generating content, so that the focus can be on optimizing the quality of output content.

[0003] However, once it is converted to a telephone voice interaction scenario, the situation is completely different. Voice interaction has extremely high real-time requirements. Even if the reply is delayed by only one or two seconds, it may cause the conversation to terminate or become unsmooth, seriously damaging the interaction effect and user experience. Although the current intelligent telephone outbound call system on the market can quickly generate replies to ensure basic efficiency, the accuracy, completeness and intelligence of the replies are greatly reduced while the architecture and computing complexity are reduced. It is difficult to reach the level of text interaction optimization, which greatly limits the in-depth application and expansion in the field of voice communication. Innovative technical solutions are urgently needed to balance the contradiction between real-time and performance in the voice interaction process, so as to promote the further development of voice interaction in this field. Summary of the invention

[0004] The embodiments of the present invention provide an intelligent telephone outbound calling method, device, electronic device and storage medium for target job recruitment to solve the problems of poor response efficiency and low accuracy in job recruitment in a telephone voice interaction scenario.

[0005] In a first aspect, an embodiment of the present invention provides an intelligent telephone outbound calling method for recruiting target positions, the method comprising:

[0006] Determine a target question content associated with a target terminal device, the target question content includes text content generated by converting a target voice input content, the target voice input content includes a voice question related to target job recruitment obtained from the target terminal device after making an outbound call to the target terminal device;

[0007] Detect whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer content for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold;

[0008] If there is a standard answer content that matches the target question content, the standard answer content that matches the target question content is determined as the target question and answer result of the target question content;

[0009] If there is no standard answer content matching the target question content, then search for target corpus related to the target question content from the preset corpus, which stores the knowledge resources needed to assist in outputting the question and answer results corresponding to the question content when performing the question and answer task;

[0010] Generate target prompt word content based on target corpus and target question content, and output target question and answer results of target question content through a preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform question and answer tasks according to question content and give question and answer results;

[0011] The target question and answer result based on the target question content is converted into voice and returned to the target terminal device.

[0012] In a second aspect, an embodiment of the present invention further provides an intelligent telephone outbound calling device for recruiting target positions, the device comprising:

[0013] A target question content determination module is used to determine the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, and the target voice input content includes voice questions related to the target position recruitment obtained from the target terminal device after making an outbound call to the target terminal device;

[0014] A standard answer content detection module is used to detect whether there is a standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold;

[0015] A target question and answer result determination module is used to determine the standard answer content matching the target question content as the target question and answer result of the target question content if there is a standard answer content matching the target question content;

[0016] A target corpus determination module is used to search for target corpus related to the target question content from a preset corpus if there is no standard answer content matching the target question content. The preset corpus stores knowledge resources required to assist in outputting the question and answer results corresponding to the question content when performing the question and answer task;

[0017] A target question-answering result output module is used to generate target prompt word content based on the target corpus and the target question content, and output the target question-answering result of the target question content through a preset question-answering model according to the target prompt word content. The preset question-answering model is a natural language model used to perform question-answering tasks according to the question content and give question-answering results;

[0018] The target question and answer result conversion module is used to convert the target question and answer result based on the target question content into voice and return it to the target terminal device.

[0019] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0020] at least one processor; and

[0021] a memory communicatively connected to at least one processor; wherein,

[0022] The memory stores a computer program that can be executed by at least one processor, and the computer program is executed by at least one processor so that the at least one processor can execute the smart phone outbound calling method for target position recruitment in any of the above embodiments.

[0023] In a fourth aspect, a computer-readable medium is also provided in an embodiment of the present invention, wherein the computer-readable medium stores computer instructions, and the computer instructions are used to enable a processor to implement the intelligent telephone outbound calling method for recruiting a target position in any of the above embodiments when executed.

[0024] The technical solution of the embodiment of the present invention determines the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, the target voice input content includes voice questions related to the recruitment of the target position obtained from the target terminal device after making an outbound call to the target terminal device, and realizes text conversion of the voice questions output by the target terminal device; detects whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, the similarity between the preset question contents corresponding to the standard answer contents in the preset knowledge repository and the target question contents is greater than a preset similarity threshold, and realizes priority judgment on whether there is standard answer content matching the target question content in the preset knowledge repository; if there is standard answer content matching the target question content, the standard answer content matching the target question content is determined as the target question and answer result of the target question content, and realizes that when there is standard answer content matching the target question content in the preset knowledge repository, the standard answer content matching the target question content is determined as the target question and answer result of the target question content. The target question and answer result of the target question content is taken as the target question content, thereby improving the efficiency of determining the target question and answer result; if there is no standard answer content matching the target question content, the target corpus related to the target question content is searched from the preset corpus, and the preset corpus stores the knowledge resources required for assisting in outputting the question and answer result corresponding to the question content when performing the question and answer task, so that when there is no standard answer content matching the target question content in the preset knowledge storage library, the target corpus that can assist in answering the target question content is searched from the preset corpus; the target prompt word content is generated based on the target corpus and the target question content, and the target question and answer result of the target question content is output through the preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform the question and answer task according to the question content and give the question and answer result, so as to ensure the quality of the target question and answer result, thereby improving the accuracy of the reply; the target question and answer result based on the target question content is converted into voice and returned to the target terminal device, so as to realize the conversion of the target question and answer result into voice and return it to the target terminal device, effectively ensuring the timeliness of the dialogue process.

[0025] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0027] Figure 1 A flowchart of a smart phone outbound calling method for target job recruitment provided by an embodiment of the present invention;

[0028] Figure 2 A flowchart of a reference corpus acquisition method provided by an embodiment of the present invention;

[0029] Figure 3 A flowchart of another smart phone outbound calling method for recruiting target positions provided by an embodiment of the present invention;

[0030] Figure 4 A schematic diagram of a scenario type in which a target question content has multiple follow-up questions during the questioning process provided by an embodiment of the present invention;

[0031] Figure 5 A structural block diagram of an intelligent telephone outbound calling device for recruiting target positions provided by an embodiment of the present invention;

[0032] Figure 6 A structural block diagram of an electronic device for implementing a smart phone outbound calling method for target job recruitment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0033] Embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not intended to limit the scope of protection of the present invention.

[0034] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0035] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0036] It should be noted that the concepts such as "first" and "second" mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0037] It should be noted that the modifications of "one" and "plurality" mentioned in the present invention are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0038] It is understandable that before using the technical solutions disclosed in the embodiments of the present invention, the type, scope of use, usage scenarios, etc. of the personal information involved in the present invention should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0039] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present invention according to the prompt message.

[0040] It is understandable that the above notification and user authorization process is only illustrative and does not constitute a limitation on the implementation of the present invention. Other methods that meet relevant laws and regulations may also be applied to the implementation of the present invention.

[0041] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0042] In the following embodiments, each embodiment provides optional features and examples at the same time, and the various features recorded in the embodiments can be combined to form multiple optional solutions, and each numbered embodiment should not be regarded as only one technical solution. In addition, the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0043] Figure 1A flow chart of a smart phone outbound calling method for target job recruitment provided in an embodiment of the present invention. The technical solution of the embodiment of the present invention is applicable to the situation of job recruitment in a smart phone outbound calling scenario. The method can be executed by a smart phone outbound calling device for target job recruitment. The smart phone outbound calling device for target job recruitment can be implemented in the form of software and / or hardware, and can be integrated on any electronic device with network communication function, which can be a mobile terminal, PC or server, etc.

[0044] like Figure 1 As shown, the smart phone outbound calling method for target position recruitment in the embodiment of the present invention may include the following process:

[0045] S101. Determine target question content associated with a target terminal device, where the target question content includes text content generated by converting target voice input content, and the target voice input content includes voice questions related to target job recruitment obtained from the target terminal device after making an outbound call to the target terminal device.

[0046] In an embodiment of the present invention, the target terminal device may refer to a device used by the applicant for the target position during voice interaction with the applicant for the target position after the target position recruitment smart phone calls outbound calls, such as a mobile phone, a landline, etc. The target voice input content may refer to voice questions related to the target position recruitment asked by the target position applicant output by the target terminal device. The target question content may refer to text questions related to the target position recruitment, which can be obtained by text conversion of the target voice input content. Exemplarily, the target question content can be generated by converting the target voice input content into text through a speech-to-text model, wherein the speech-to-text model can be an automatic speech recognition (ASR) model. Target position recruitment may refer to the process of recruiting objects with specific conditions to fill target position vacancies through outbound calls, such as the recruitment of drivers for online car-hailing.

[0047] Specifically, after making an outbound call on a smart phone for recruitment of a target position, voice questions related to the recruitment of the target position asked by the target position applicant can be obtained from the target terminal device as the target voice input content. Furthermore, the text questions generated after the target voice input content is converted into text can be used as the target question content associated with the target terminal device. Exemplarily, the target question content can be "What is the specific job content of the target position? What are the working hours of the target position?".

[0048] S102. Detect whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository being used to record a number of preset question contents and standard answer content for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold.

[0049] In an embodiment of the present invention, the preset question content may refer to a preset frequently asked question related to the target position recruitment, such as a question for a conceptual answer. The standard answer content may refer to a preset standard answer corresponding to the preset question content. The preset knowledge repository may refer to a preset knowledge repository storing preset question content and standard answer content corresponding to the preset question content. Exemplarily, the preset knowledge repository may be a Frequently Asked Questions (FAQ) library related to the target position recruitment.

[0050] Specifically, the similarity comparison result can be obtained by performing a similarity comparison between the target question content and several preset question contents in the preset knowledge repository. Exemplarily, the target question content and several preset question contents can be compared for similarity in sequence or simultaneously. Furthermore, it can be determined based on the similarity comparison result whether there is a preset question content in the preset knowledge repository whose similarity with the target question content is greater than a preset similarity threshold, thereby determining whether there is a standard answer content matching the target question content in the preset knowledge repository.

[0051] As an optional but non-limiting implementation method, detecting whether there is standard answer content matching the target question content in the preset knowledge repository includes the following steps A1-A3:

[0052] Step A1, searching for candidate answer content from a preset knowledge repository, wherein the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is greater than the similarity between the preset question content corresponding to the standard answer content in the preset knowledge repository except the candidate answer content and the target question content.

[0053] Step A2: If the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is greater than a first preset similarity, the candidate answer content is determined as the standard answer content matching the target question content.

[0054] Step A3: If the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is not greater than the first preset similarity, it is determined that there is no standard answer content matching the target question content in the preset knowledge repository.

[0055] In an embodiment of the present invention, the candidate answer content may refer to the standard answer content corresponding to the preset question content in the preset knowledge repository that has the greatest similarity to the target question content. The first preset similarity may refer to a preset similarity threshold between the target question content and the preset question content in the preset knowledge repository, which is used to determine whether there is a standard answer content matching the target question content in the preset knowledge repository.

[0056] Specifically, after comparing the similarity between the target question content and several preset question contents in the preset knowledge repository, the preset question content with the greatest similarity to the target question content in the preset knowledge repository can be used as the candidate question content, so that the standard answer content corresponding to the candidate question content can be used as the candidate answer content. Furthermore, the similarity between the target question content and the candidate question content can be compared with the first preset similarity. If the similarity between the target question content and the candidate question content is greater than the first preset similarity, the candidate answer content can be used as the standard answer content that matches the target question content. If the similarity between the target question content and the candidate question content is not greater than the first preset similarity, it can be considered that there is no standard answer content that matches the target question content in the preset knowledge repository.

[0057] S103. If there is standard answer content that matches the target question content, the standard answer content that matches the target question content is determined as the target question and answer result of the target question content.

[0058] Specifically, if there is a preset question content in the preset knowledge repository whose similarity with the target question content is greater than a preset similarity threshold, the standard answer content corresponding to the preset question content in the preset knowledge repository whose similarity with the target question content is greater than the preset similarity threshold can be used as the standard answer content matching the target question content. Furthermore, the target question content matching the standard answer content can be used as the target question and answer result of the target question content.

[0059] S104. If there is no standard answer content that matches the target question content, search for target corpus related to the target question content from a preset corpus, which stores knowledge resources required to assist in outputting question and answer results corresponding to the question content when performing question and answer tasks.

[0060] In an embodiment of the present invention, the preset corpus may refer to a corpus that is pre-set and stores knowledge resources required for assisting in outputting target question content corresponding to target question and answer results. Exemplarily, the preset corpus may be a Retrieval Augmented Generation (RAG) corpus that matches the target question content.

[0061] Specifically, if there is no preset question content in the preset knowledge repository whose similarity with the target question content is greater than the preset similarity threshold, it can be considered that there is no standard answer content matching the target question content in the preset knowledge repository. Furthermore, relevant knowledge resources that assist in answering the target question content can be searched from the preset corpus as the target corpus.

[0062] S105. Generate target prompt word content based on the target corpus and the target question content, and output the target question and answer result of the target question content through a preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform question and answer tasks according to the question content and give question and answer results.

[0063] In an embodiment of the present invention, a natural language model may refer to a machine learning model that can understand and generate natural language text, and is used to process and analyze large amounts of language data to perform various tasks. A preset question-and-answer model may refer to a natural language model that is generated in advance through model fine-tuning training and can give corresponding question-and-answer results based on the content of the question. Exemplarily, in order to improve the response efficiency of the preset question-and-answer model, a model deployed locally may be used for fine-tuning training to generate a preset question-and-answer model.

[0064] Specifically, the target corpus and the target question content can be combined to form the target prompt word content and input into the preset question-answering model, so as to use the preset question-answering model to output the corresponding text reply content and serve as the target question-answering result of the target question content.

[0065] As an optional but non-limiting implementation method, the model building process of the preset question-answering model includes the following steps B1-B2:

[0066] Step B1, determine multiple reference corpora for fine-tuning the question-answering model. The multiple reference corpora are reference question contents generated by extracting the text content of historical outbound call recording data, and question-answer pairs consisting of different reference question-answer results attached to the reference question contents. The numerical information in the reference corpora is pre-masked.

[0067] Step B2: fine-tune the question-answering model to be trained based on multiple reference corpora to obtain a preset question-answering model.

[0068] In an embodiment of the present invention, the reference question content may refer to the question corresponding to the outbound call recording data obtained by injecting the relevant information extracted from the historical outbound call recording data after text conversion into the system prompt template. The reference question and answer result may refer to the reference reply to the reference question content. The reference corpus may refer to the reference question content used for fine-tuning training of the preset question and answer model and the question and answer pairs composed of different reference question and answer results corresponding to the reference question content.

[0069] Exemplary, reference Figure 2 , the relevant information extracted from the historical outbound call recording data after text conversion can be used as training corpus, and multiple text blocks can be obtained after slicing the training corpus. Questions and corresponding answers extracted from multiple text blocks can constitute question-answer pairs, and similarity deduplication of question-answer pairs can be performed to obtain deduplicated question-answer pairs to remove similar question-answer pairs. Furthermore, in order to prevent the preset question-answer model from directly relying on the numerical information contained in the reference corpus, and in order to enable the preset question-answer model to generate responses based on the information injected into the reference corpus, all numerical information in the deduplicated question-answer pairs is pre-masked to obtain the reference corpus, for example, all numerical information is converted into special symbols.

[0070] Specifically, multiple reference corpora can be established based on historical outbound call recording data, including reference question content and different reference question and answer results corresponding to the reference question content to form question and answer pairs. Then, the preset question and answer model is obtained by fine-tuning the to-be-trained question and answer model through multiple reference corpora. Exemplarily, the preset question and answer model can be obtained through supervised fine tuning (SFT) training.

[0071] Optionally, a model evaluation is performed on the preset question-and-answer model based on preset measurement model indicators and simulated conversation records. The preset measurement model indicators may include at least one of: rationality, reference, fluency, non-repetitiveness, consistency, and effectiveness. Among them, rationality may refer to measuring whether the results output by the preset question-and-answer model meet the actual conditions; reference may refer to measuring the reference effect of the preset question-and-answer model on the information injected into the reference corpus; fluency may refer to measuring whether the reply content of the preset question-and-answer model meets the semantic specifications of the target position; non-repetitiveness may refer to measuring whether the similarity of the reply content of the preset question-and-answer model cannot exceed the preset threshold; consistency may refer to measuring the topic consistency between the reply content of the preset question-and-answer model and the target question content; effectiveness may refer to measuring whether the reply content of the preset question-and-answer model can provide effective information to prevent the preset question-and-answer model from always avoiding answers. The simulated conversation record may refer to the conversation record generated after simulating a conversation between applicants with different conditions and the preset question-and-answer model.

[0072] S106: The target question and answer result based on the target question content is converted into voice and returned to the target terminal device.

[0073] Specifically, the target question and answer result of the target question content is converted into a voice and the generated voice question and answer result is returned to the target terminal device, so as to realize the smart phone outbound call to the target terminal device. Exemplarily, the voice question and answer result can be generated by converting the target question and answer result into a voice through a text to speech (TTS) model.

[0074] The technical solution of the embodiment of the present invention determines the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, the target voice input content includes voice questions related to the recruitment of the target position obtained from the target terminal device after making an outbound call to the target terminal device, and realizes text conversion of the voice questions output by the target terminal device; detects whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, the similarity between the preset question contents corresponding to the standard answer contents in the preset knowledge repository and the target question contents is greater than a preset similarity threshold, and realizes priority judgment on whether there is standard answer content matching the target question content in the preset knowledge repository; if there is standard answer content matching the target question content, the standard answer content matching the target question content is determined as the target question and answer result of the target question content, and realizes that when there is standard answer content matching the target question content in the preset knowledge repository, the standard answer content matching the target question content is determined as the target question and answer result of the target question content. The target question and answer result of the target question content is taken as the target question content, thereby improving the efficiency of determining the target question and answer result; if there is no standard answer content matching the target question content, the target corpus related to the target question content is searched from the preset corpus, and the preset corpus stores the knowledge resources required for assisting in outputting the question and answer result corresponding to the question content when performing the question and answer task, so that when there is no standard answer content matching the target question content in the preset knowledge storage library, the target corpus that can assist in answering the target question content is searched from the preset corpus; the target prompt word content is generated based on the target corpus and the target question content, and the target question and answer result of the target question content is output through the preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform the question and answer task according to the question content and give the question and answer result, so as to ensure the quality of the target question and answer result, thereby improving the accuracy of the reply; the target question and answer result based on the target question content is converted into voice and returned to the target terminal device, so as to realize the conversion of the target question and answer result into voice and return it to the target terminal device, effectively ensuring the timeliness of the dialogue process.

[0075] Figure 3 A flowchart of another smart phone outbound calling method for recruiting target positions provided in an embodiment of the present invention. The technical solution of this embodiment further optimizes the process of searching for target corpus related to the target question content from a preset corpus in the aforementioned embodiment on the basis of the technical solution of the aforementioned embodiment. This embodiment can be combined with various optional solutions in one or more of the aforementioned embodiments.

[0076] like Figure 3As shown, the smart phone outbound calling method for target position recruitment in the embodiment of the present invention may include the following process:

[0077] S201. Determine target question content associated with a target terminal device, where the target question content includes text content generated by converting target voice input content, and the target voice input content includes voice questions related to target job recruitment obtained from the target terminal device after an outbound call is made to the target terminal device.

[0078] S202. Detect whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository being used to record a number of preset question contents and standard answer content for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold.

[0079] S203: If there is standard answer content that matches the target question content, the standard answer content that matches the target question content is determined as the target question and answer result of the target question content.

[0080] S204. If there is no standard answer content that matches the target question content, determine the expected type of question and answer result corresponding to the target question content. The expected type of question and answer result is determined based on the level of detail of the question and answer result expected for the question content.

[0081] In the embodiment of the present invention, the expected type of question and answer result may refer to the scenario type of the target question content asking information. Different expected types of question and answer results require different levels of detail of the response content corresponding to the target question content.

[0082] Specifically, if there is no standard answer content matching the target question content in the preset knowledge repository, the scene recognition is performed based on the target question content to determine the query information of the target question content. Further, the expected level of detail of the question and answer result is determined based on the importance of the query information of the target question content, and the expected type of the question and answer result corresponding to the target question content is further determined based on the expected level of detail of the question and answer result.

[0083] S205. According to the expected type of question and answer result corresponding to the target question content, a target corpus related to the target question content is searched from a preset corpus. The level of detail of the knowledge resources provided by the target corpus when answering the target question content is correlated with the expected type of question and answer result corresponding to the target question content.

[0084] In an embodiment of the present invention, the degree of detail of the knowledge resources provided by the target corpus for the target question content is related to the degree of detail of the expected question and answer result indicated by the expected type of the question and answer result corresponding to the target question content. The greater the degree of detail of the expected question and answer result indicated by the expected type of the question and answer result corresponding to the target question content, the greater the degree of detail of the knowledge resources provided by the target corpus for the target question content.

[0085] Specifically, the level of detail of the question and answer results can be given according to the expectations expressed by the expected type of question and answer results corresponding to the target question content, and the target corpus can be searched from the preset corpus to provide the knowledge resources with the corresponding level of detail required for answering the target question content.

[0086] As an optional but non-limiting implementation method, the expected type of question-answering result includes a scenario type in which there are multiple follow-up questions in the question-asking process according to the target question content and a scenario type in which the question-answering result needs to be fuzzy processed in the question-asking process according to the target question content; accordingly, according to the expected type of question-answering result corresponding to the target question content, searching for target corpus related to the target question content from a preset corpus includes the following steps C1-C2:

[0087] Step C1. If the expected type of question and answer result corresponding to the target question content indicates a scenario type in which there are multiple follow-up questions in the questioning process according to the target question content, then the number of follow-up questions in the questioning process according to the target question content is determined based on the question and answer dialogue record associated with the target terminal device, and the target corpus related to the target question content is searched from the preset corpus according to the number of follow-up questions in the questioning process according to the target question content; wherein, the more follow-up questions are asked in the questioning process according to the target question content, the more detailed the knowledge resources provided by the target corpus when asking and answering the target question content.

[0088] Step C2: If the expected type of question and answer result corresponding to the target question content indicates a scenario type in which the question and answer result needs to be fuzzified during the questioning process according to the target question content, then a target corpus related to the target question content is searched from a plurality of fuzzified corpora stored in a preset corpus.

[0089] In the embodiment of the present invention, the scene type of multiple questions indicates that the level of detail of the expected question and answer results is related to the number of questions asked, and accordingly, the level of detail of the knowledge resources provided by the target corpus is also related to the number of questions asked. The scene type of fuzzy processing of the question and answer results indicates that the question and answer results given only need to be briefly introduced.

[0090] Specifically, if the expected type of question-answering result corresponding to the target question content indicates a scenario type in which the target question content has multiple follow-up questions during the questioning process, the number of follow-up questions of the target question content during the questioning process is recorded. For example, the number of follow-up questions can be stored in a configuration information cache table. Furthermore, according to the number of follow-up questions of the target question content, a target corpus that is related to the target question content and meets the knowledge resource with a level of detail corresponding to the number of follow-up questions can be searched from a preset corpus.

[0091] Exemplary, reference Figure 4 When the target question involves a specific value, if a fuzzy response strategy is used, follow-up questions will be asked, and the content can be gradually expanded in detail. When the question is asked for the first time, the preset corpus 1 in the preset corpus is used to provide reference words and information injection, and the number of follow-up questions is updated to 1, for example, "the value is between 2000-3000". When the question is asked again, the preset corpus 2 in the preset corpus is used to provide reference words and information injection, and the number of follow-up questions is updated to 2, for example, "the value is about 2500". When the number of follow-up questions exceeds 2, the preset corpus 3 in the preset corpus is used to provide reference words and information injection. Among them, the degree of detail of preset corpus 1, preset corpus 2 to preset corpus 3 increases in sequence.

[0092] If the expected type of question and answer result corresponding to the target question content indicates the scenario type in which the question and answer result of the target question content needs to be fuzzified during the questioning process, multiple target corpora that have been fuzzified and are related to the target question content are searched from the preset corpus to avoid providing the question and answer result corresponding to the target question content in full at one time. Exemplarily, the target corpora stored in the preset corpus can be transcribed to obtain the fuzzified corpora.

[0093] S206. Generate target prompt word content based on the target corpus and the target question content, and output the target question and answer result of the target question content through a preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform question and answer tasks according to the question content and give question and answer results.

[0094] S207: The target question and answer result based on the target question content is converted into voice and returned to the target terminal device.

[0095] As an optional but non-limiting implementation, the target question and answer result based on the target question content is converted into voice and returned to the target terminal device, including the following steps D1-D3:

[0096] Step D1: After the target question and answer result of the target question content is output through the preset question and answer model, additional question content associated with the target question content is extracted from the preset question configuration information, the additional question content includes required question content and / or optional question content, and the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is judged.

[0097] Step D2: If the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is less than a second preset similarity, speech conversion is performed based on the additional question content associated with the target question content and the target question and answer result of the target question content and returned to the target terminal device.

[0098] Step D3: If the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is not less than a second preset similarity, the target question and answer result of the target question content is converted into speech and returned to the target terminal device.

[0099] In an embodiment of the present invention, additional question content may refer to additional questions asked when returning the target question and answer results to the target terminal device after the target position recruitment smart phone calls out. The additional question content includes mandatory question content, optional question content, or mandatory question content and optional question content. The mandatory question content may refer to content such as the recruitment requirements for the target position. The optional question content may refer to questions defined according to the business. The priority of the mandatory question content is greater than the optional question content, and both the mandatory question content and the optional question content can be used to determine the willingness to apply for the target position. The preset question configuration information may refer to pre-set configuration information that stores the additional question content. Exemplarily, the preset question configuration information may be stored in the configuration information cache table.

[0100] Specifically, after the target question and answer result of the target question content is output through the preset question and answer model, the additional question content associated with the target question content can be extracted from the preset question configuration information. Furthermore, the additional question content associated with the target question content can be compared for similarity with the target question and answer result of the target question content output by the preset question and answer model. If the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is not less than the second preset similarity, it means that the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model do not need to be combined, and the target question and answer result of the target question content output by the preset question and answer model only needs to be returned to the target terminal device after voice conversion, so as to avoid the target question and answer result output by the preset question and answer model being the same as the additional question content associated with the target question content, thereby causing repeated inquiries.

[0101] If the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is less than a second preset similarity, the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model are combined and returned to the target terminal device through voice conversion, and the additional question content in the preset question configuration information is updated in a timely manner to prevent repeated questions. Exemplarily, the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model can be integrated through a preset base model. The preset base model does not need to be fine-tuned. It only needs to accept the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model through pre-designed relevant prompt words, and integrate them into complete content for output.

[0102] As an optional but non-limiting implementation method, the intelligent telephone outbound calling method for target position recruitment in an embodiment of the present invention also includes: combining the target question and answer results of the target question content output each time to generate prompt words and inputting them into the dialogue summary model, and outputting the target position recruitment results through the dialogue summary model. The target position recruitment results are used to reflect the possibility of job hunting and applying for the target position.

[0103] In an embodiment of the present invention, the target position recruitment result may refer to the degree of intention to seek a job or apply for the target position. The target position recruitment result may include positions that can be discussed and positions that cannot be discussed. Among them, positions that can be discussed may include clear intentions and wavering intentions. Clear intentions may refer to clear expressions of intentions to join the job or apply for the job, and further detailed communication may be possible. Wavering intentions may refer to no clear expressions of intentions to join the job or apply for the job, indicating a wait-and-see comparison. Position recruitment that cannot be discussed may include no intentions at the moment, conditions not met, and failure to establish contact. No intentions at the moment may refer to clear expressions of intentions to refuse to join the job or apply for the job. Conditions not met may refer to non-compliance with the relevant requirements for the target position recruitment.

[0104] Specifically, the target question and answer results of the target question content output each time can be updated and recorded. Exemplarily, the target question and answer results can be stored in the conversation record cache table. Furthermore, after the smart phone outbound call for the target position recruitment is completed, the target question and answer results recorded each time are summarized to generate corresponding prompt words, and the prompt words are input into the conversation summary model to obtain the corresponding target position recruitment results. Among them, the content quality of the target position recruitment results output by the conversation summary model meets the preset content quality threshold.

[0105] The technical solution of the embodiment of the present invention determines the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, the target voice input content includes voice questions related to the recruitment of the target position obtained from the target terminal device after making an outbound call to the target terminal device, and realizes text conversion of the voice questions output by the target terminal device; detects whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a plurality of preset question contents and standard answer contents for the preset question contents, the preset question contents corresponding to the standard answer contents matching the target question contents in the preset knowledge repository are matched with the target question contents The similarity between the target question content and the target answer content is greater than the preset similarity threshold, so that it is realized to prioritize whether there is standard answer content matching the target question content in the preset knowledge repository; if there is standard answer content matching the target question content, the standard answer content matching the target question content is determined as the target question and answer result of the target question content, so that when there is standard answer content matching the target question content in the preset knowledge repository, the standard answer content matching the target question content is used as the target question and answer result of the target question content, thereby improving the efficiency of determining the target question and answer result; if there is no standard answer content matching the target question content, the expected type of the question and answer result corresponding to the target question content is determined, and the expected type of the question and answer result is determined according to the target question content. The expected level of detail of the question and answer results given by the question content is determined, so that when there is no standard reply content matching the target question content in the preset knowledge repository, the expected type of question and answer results corresponding to the target question content can be determined according to the expected level of detail of the question and answer results given by the target question content; according to the expected type of question and answer results corresponding to the target question content, the target corpus related to the target question content is searched from the preset corpus, and the level of detail of the knowledge resources provided by the target corpus when answering the target question content is related to the expected type of question and answer results corresponding to the target question content, so that the target corpus that can assist in answering the target question content can be searched from the preset corpus according to the expected type of question and answer results corresponding to the target question content. corpus, so that the generated target question and answer result can accurately match the target question content; generate the target prompt word content based on the target corpus and the target question content, and output the target question and answer result of the target question content through the preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform question and answer tasks according to the question content and give question and answer results. It realizes the use of the preset question and answer model to generate the target question and answer result of the target question content, so as to ensure the quality of the target question and answer result, thereby improving the accuracy of the reply; the target question and answer result based on the target question content is converted into speech and returned to the target terminal device, realizing the conversion of the target question and answer result into speech and returning it to the target terminal device, effectively ensuring the timeliness of the dialogue process.

[0106] Figure 5 The present invention provides a structural block diagram of a smart phone outbound calling device for recruiting target positions in an embodiment of the present invention. The technical solution of the embodiment of the present invention is applicable to the situation of recruiting positions in a smart phone outbound calling scenario. The smart phone outbound calling device for recruiting target positions can be implemented in the form of software and / or hardware, and is generally integrated on any electronic device with network communication function, which can be a mobile terminal, PC or server, etc.

[0107] like Figure 5 As shown, the smart phone outbound calling device for target job recruitment in the embodiment of the present invention may include the following:

[0108] The target question content determination module 301 is used to determine the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, and the target voice input content includes voice questions related to the target position recruitment obtained from the target terminal device after making an outbound call to the target terminal device;

[0109] The standard answer content detection module 302 is used to detect whether there is a standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold;

[0110] A target question and answer result determination module 303 is used to determine the standard answer content matching the target question content as the target question and answer result of the target question content if there is a standard answer content matching the target question content;

[0111] The target corpus determination module 304 is used to search for target corpus related to the target question content from a preset corpus if there is no standard answer content matching the target question content, wherein the preset corpus stores knowledge resources required for assisting in outputting question and answer results corresponding to the question content when performing the question and answer task;

[0112] A target question-answering result output module 305 is used to generate target prompt word content based on the target corpus and the target question content, and output the target question-answering result of the target question content through a preset question-answering model according to the target prompt word content. The preset question-answering model is a natural language model used to perform a question-answering task according to the question content and give a question-answering result;

[0113] The target question and answer result conversion module 306 is used to convert the target question and answer result based on the target question content into voice and return it to the target terminal device.

[0114] On the basis of any of the above optional technical solutions, optionally, the standard answer content detection module 302 includes: a candidate answer content determination unit and a standard answer content determination unit. The candidate answer content determination unit is used to search for candidate answer content from the preset knowledge repository, and the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is greater than the similarity between the preset question content corresponding to the standard answer content in the preset knowledge repository other than the candidate answer content and the target question content in the preset knowledge repository; the standard answer content determination unit is used to determine the candidate answer content as the standard answer content matching the target question content if the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is greater than the first preset similarity; the standard answer content determination unit is also used to determine that there is no standard answer content matching the target question content in the preset knowledge repository if the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is not greater than the first preset similarity.

[0115] On the basis of any of the above optional technical solutions, optionally, the target corpus determination module 304 includes: a question and answer result expected type determination unit and a target corpus search unit. Among them, the question and answer result expected type determination unit is used to determine the expected type of question and answer result corresponding to the target question content, and the expected type of question and answer result is determined according to the level of detail of the question and answer result expected for the question content; the target corpus search unit is used to search for the target corpus related to the target question content from the preset corpus according to the expected type of question and answer result corresponding to the target question content, and the level of detail of the knowledge resources provided by the target corpus when answering the target question content is related to the expected type of question and answer result corresponding to the target question content.

[0116] Based on any of the above optional technical solutions, optionally, the expected types of question and answer results include scenario types in which there are multiple follow-up questions in the questioning process according to the target question content and scenario types in which the question and answer results need to be fuzzy processed in the questioning process according to the target question content; accordingly, the target corpus search unit includes: a first search sub-unit and a second search sub-unit. Among them, the first search sub-unit is used to determine the number of follow-up questions in the questioning process according to the target question content based on the question and answer dialogue record associated with the target terminal device if the expected type of question and answer result corresponding to the target question content indicates a scenario type in which there are multiple follow-up questions in the questioning process according to the target question content, and search for the target corpus related to the target question content from the preset corpus according to the number of follow-up questions in the questioning process according to the target question content; wherein, the more follow-up questions are asked in the questioning process according to the target question content, the more detailed the knowledge resources provided by the target corpus when asking and answering the target question content; the second search sub-unit is used to search for the target corpus related to the target question content from multiple fuzzy corpora stored in the preset corpus if the expected type of question and answer result corresponding to the target question content indicates a scenario type in which the question and answer result needs to be fuzzified in the questioning process according to the target question content.

[0117] Based on any of the above optional technical solutions, optionally, the model building process of the preset question and answer model includes: determining multiple reference corpora for fine-tuning the question and answer model, the multiple reference corpora are reference question contents generated by extracting the text content of conversation recordings of historical outbound call recordings and question and answer pairs composed of different reference question and answer results attached to the reference question contents, and the numerical information in the reference corpora is masked in advance; fine-tuning the question and answer model to be trained based on the multiple reference corpora to obtain the preset question and answer model.

[0118] Based on any of the above optional technical solutions, optionally, the target question and answer result conversion module 306 includes: an additional question content determination unit and a target question and answer result determination unit. Among them, the additional question content determination unit is used to extract the additional question content associated with the target question content from the preset question configuration information after the target question and answer result of the target question content is output through the preset question and answer model, the additional question content includes required question content and / or optional question content, and the additional question content associated with the target question content is judged for similarity with the target question and answer result of the target question content output by the preset question and answer model; the target question and answer result determination unit is used to perform voice conversion based on the additional question content associated with the target question content and the target question and answer result of the target question content if the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is less than a second preset similarity, and return it to the target terminal device; the target question and answer result determination unit is also used to perform voice conversion on the target question content and return it to the target terminal device if the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is not less than a second preset similarity.

[0119] Based on any of the above optional technical solutions, the target job recruitment smart phone outbound calling device of the embodiment of the present invention may further include: a target job recruitment result determination module. The target job recruitment result determination module is used to combine the target question and answer results of the target question content output each time to generate prompt words and input them into the dialogue summary model, and output the target job recruitment result through the dialogue summary model. The target job recruitment result is used to reflect the possibility of job hunting and applying for the target job.

[0120] The technical solution of the embodiment of the present invention is to determine the target question content associated with the target terminal device through the target question content determination module 301, the target question content includes the text content generated by converting the target voice input content, and the target voice input content includes the voice question related to the target position recruitment obtained from the target terminal device after making an outbound call to the target terminal device, so as to realize the text conversion of the voice question output by the target terminal device; to detect whether there is a standard reply content matching the target question content in the preset knowledge repository through the standard reply content detection module 302, the preset knowledge repository is used to record a plurality of preset question contents and standard reply contents for the preset question contents, and the similarity between the preset question contents corresponding to the standard reply contents matching the target question contents in the preset knowledge repository and the target question contents is greater than the preset similarity threshold, so as to realize the priority judgment whether there is a standard reply content matching the target question content in the preset knowledge repository; to determine through the target question and answer result determination module 303 if there is a standard reply content matching the target question content, then the standard reply content matching the target question content is determined as the target question and answer result of the target question content, so as to realize that when there is a standard reply content matching the target question content in the preset knowledge repository, the standard reply content matching the target question content is converted into the target question and answer result. The target question content is used as the target question and answer result of the target question content, thereby improving the efficiency of determining the target question and answer result; if it is determined through the target corpus determination module 304 that there is no standard answer content matching the target question content, then the target corpus related to the target question content is searched from the preset corpus, and the preset corpus stores the knowledge resources required for assisting in outputting the question and answer result corresponding to the question content when performing the question and answer task, so that when there is no standard answer content matching the target question content in the preset knowledge storage library, the target corpus that can assist in answering the target question content is searched from the preset corpus; the target question and answer result output module 305 determines the target corpus based on the target corpus and the target The target question content generates the target prompt word content, and the target question and answer result of the target question content is output through the preset question and answer model according to the target prompt word content. The preset question and answer model is a natural language model used to perform question and answer tasks according to the question content and give question and answer results. It realizes the use of the preset question and answer model to generate the target question and answer result of the target question content, so as to ensure the quality of the target question and answer result, thereby improving the accuracy of the reply; the target question and answer result conversion module 306 performs voice conversion based on the target question and answer result of the target question content and returns it to the target terminal device, thereby realizing the conversion of the target question and answer result into voice and returning it to the target terminal device, effectively ensuring the timeliness of the dialogue process.

[0121] The names of the messages or information exchanged between multiple devices in the embodiments of the present invention are only used for illustrative purposes, and are not used to limit the scope of these messages or information.

[0122] The smart phone outbound calling device for target position recruitment provided by the embodiment of the present invention can execute the smart phone outbound calling method for target position recruitment provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the smart phone outbound calling method for target position recruitment.

[0123] It is worth noting that the various units and modules included in the above-mentioned device are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the embodiments of the present invention.

[0124] Figure 6 The structure block diagram of an electronic device for implementing a smart phone outbound calling method for target job recruitment provided by an embodiment of the present invention. Figure 6 , which shows an electronic device (eg Figure 6 The terminal device in the embodiment of the present invention may include but is not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0125] like Figure 6 As shown, the electronic device 400 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage device 406 into a random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the electronic device 400 are also stored. The processing device 401, the ROM 402, and the RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0126] Typically, the following devices may be connected to the I / O interface 405: an input device 406 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 407 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 406 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 409. The communication device 409 may allow the electronic device 400 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 6The electronic device 400 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0127] The names of the messages or information exchanged between multiple devices in the embodiments of the present invention are only used for illustrative purposes, and are not used to limit the scope of these messages or information.

[0128] The electronic device provided in the embodiment of the present invention and the smart phone outbound calling method for target position recruitment provided in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0129] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains a program code for executing the smart phone outbound calling method for target position recruitment shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 409, or installed from the storage device 406, or installed from the ROM 402. When the computer program is executed by the processing device 401, the above-mentioned functions defined in the smart phone outbound calling method for target position recruitment of the embodiment of the present invention are executed.

[0130] An embodiment of the present invention provides a computer storage medium on which a computer program is stored. When the program is executed by a processor, the intelligent telephone outbound calling method for recruiting target positions provided in the above embodiment is implemented.

[0131] It should be noted that the computer-readable medium of the present invention can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0132] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0133] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0134] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: determines the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, and the target voice input content includes voice questions related to the recruitment of the target position obtained from the target terminal device after making an outbound call to the target terminal device; detects whether there is standard answer content matching the target question content in the preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than the preset similarity degree threshold; if there is standard reply content that matches the target question content, the standard reply content that matches the target question content is determined as the target question and answer result of the target question content; if there is no standard reply content that matches the target question content, the target corpus related to the target question content is searched from the preset corpus, and the preset corpus stores the knowledge resources required for assisting in outputting the question and answer results corresponding to the question content when performing the question and answer task; the target prompt word content is generated based on the target corpus and the target question content, and the target question and answer result of the target question content is output through the preset question and answer model according to the target prompt word content, and the preset question and answer model is a natural language model used to perform the question and answer task according to the question content and give the question and answer result; the target question and answer result based on the target question content is converted into speech and returned to the target terminal device.

[0135] Alternatively, the computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device executes the smart phone outbound calling method for target position recruitment in any one of the above embodiments.

[0136] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0137] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0138] The units involved in the embodiments of the present invention may be implemented by software or hardware. The name of a unit does not limit the unit itself in some cases. For example, the first acquisition unit may also be described as a "unit for acquiring at least two Internet Protocol addresses".

[0139] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0140] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0141] The above description is only a preferred embodiment of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the disclosure scope involved in the present invention is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present invention (but not limited to) by each other.

[0142] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the present invention. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0143] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. An intelligent telephone outbound calling method for recruiting target positions, characterized in that: The method comprises: Determine a target question content associated with a target terminal device, the target question content includes text content generated by converting a target voice input content, the target voice input content includes a voice question related to target job recruitment obtained from the target terminal device after making an outbound call to the target terminal device; Detecting whether there is a standard answer content matching the target question content in a preset knowledge repository, wherein the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold; If there is no standard answer content matching the target question content, search for target corpus related to the target question content from a preset corpus, wherein the preset corpus stores knowledge resources required for assisting in outputting question and answer results corresponding to the question content when performing the question and answer task; Generate a target prompt word content based on the target corpus and the target question content, and output a target question and answer result of the target question content through a preset question and answer model according to the target prompt word content, wherein the preset question and answer model is a natural language model used to perform a question and answer task according to the question content and give a question and answer result; The target question and answer result based on the target question content is converted into voice and returned to the target terminal device.

2. The method according to claim 1, characterized in that Detecting whether there is standard answer content matching the target question content in the preset knowledge repository includes: Searching for candidate answer content from the preset knowledge repository, wherein the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is greater than the similarity between the preset question content corresponding to the standard answer content in the preset knowledge repository other than the candidate answer content and the target question content; If the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is greater than a first preset similarity, the candidate answer content is determined as the standard answer content matching the target question content; If the similarity between the preset question content corresponding to the candidate answer content in the preset knowledge repository and the target question content is not greater than the first preset similarity, it is determined that there is no standard answer content matching the target question content in the preset knowledge repository.

3. The method according to claim 1, characterized in that Searching for target corpus related to the target question content from a preset corpus, including: Determine the expected type of question and answer result corresponding to the target question content, wherein the expected type of question and answer result is determined according to the level of detail of the question and answer result expected to be given for the question content; According to the expected type of question and answer result corresponding to the target question content, a target corpus related to the target question content is searched from a preset corpus, and the degree of detail of the knowledge resources provided by the target corpus when answering the target question content is related to the expected type of question and answer result corresponding to the target question content.

4. The method according to claim 3, characterized in that The expected types of question-and-answer results include scenario types in which there are multiple follow-up questions in the question-and-answer process according to the target question content and scenario types in which the question-and-answer results need to be fuzzy processed in the question-and-answer process according to the target question content; Accordingly, according to the expected type of question and answer result corresponding to the target question content, the target corpus related to the target question content is searched from the preset corpus, including: If the expected type of question-and-answer result corresponding to the target question content indicates a scenario type in which there are multiple follow-up questions in the questioning process according to the target question content, the number of follow-up questions in the questioning process according to the target question content is determined based on the question-and-answer dialogue record associated with the target terminal device, and the target corpus related to the target question content is searched from the preset corpus according to the number of follow-up questions in the questioning process according to the target question content; wherein, the more the number of follow-up questions in the questioning process according to the target question content, the greater the degree of detail of the knowledge resources provided by the target corpus when asking and answering questions for the target question content; If the expected type of question and answer result corresponding to the target question content indicates a scenario type in which the question and answer result needs to be fuzzified during the questioning process according to the target question content, then the target corpus related to the target question content is searched from multiple fuzzified corpora stored in a preset corpus.

5. The method according to claim 1, characterized in that The model building process of the preset question-answering model includes: Determine a plurality of reference corpora for fine-tuning the question-answering model, wherein the plurality of reference corpora are reference question contents generated by extracting text contents of historical outbound call conversation recording data and question-answer pairs consisting of different reference question-answer results attached to the reference question contents, and the numerical information in the reference corpora is pre-masked; The preset question-answering model is obtained by fine-tuning the question-answering model to be trained based on the multiple reference corpora.

6. The method according to claim 1, characterized in that The target question and answer result based on the target question content is converted into voice and returned to the target terminal device, including: After the target question-answering result of the target question content is outputted by the preset question-answering model, additional question content associated with the target question content is extracted from the preset question configuration information, wherein the additional question content includes required question content and / or optional question content, and similarity is judged between the additional question content associated with the target question content and the target question-answering result of the target question content outputted by the preset question-answering model; If the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is less than a second preset similarity, then voice conversion is performed based on the additional question content associated with the target question content and the target question and answer result of the target question content and returned to the target terminal device; If the similarity between the additional question content associated with the target question content and the target question and answer result of the target question content output by the preset question and answer model is not less than a second preset similarity, the target question and answer result of the target question content is converted into speech and returned to the target terminal device.

7. The method according to claim 6, characterized in that The method further comprises: The target question and answer results of the target question content output each time are combined to generate prompt words and input into the dialogue summary model, and the target position recruitment results are output through the dialogue summary model. The target position recruitment results are used to reflect the possibility of job hunting and applying for the target position.

8. An intelligent telephone outbound calling device for recruiting target positions, characterized in that: The device comprises: A target question content determination module, used to determine the target question content associated with the target terminal device, the target question content includes text content generated by converting the target voice input content, the target voice input content includes voice questions related to the target position recruitment obtained from the target terminal device after making an outbound call to the target terminal device; A standard answer content detection module is used to detect whether there is a standard answer content matching the target question content in a preset knowledge repository, the preset knowledge repository is used to record a number of preset question contents and standard answer contents for the preset question contents, and the similarity between the preset question content corresponding to the standard answer content matching the target question content in the preset knowledge repository and the target question content is greater than a preset similarity threshold; A target corpus determination module is used to search for target corpus related to the target question content from a preset corpus if there is no standard answer content matching the target question content, wherein the preset corpus stores knowledge resources required for assisting in outputting question and answer results corresponding to the question content when performing the question and answer task; A target question-answering result output module is used to generate a target prompt word content based on the target corpus and the target question content, and output a target question-answering result of the target question content through a preset question-answering model according to the target prompt word content, wherein the preset question-answering model is a natural language model used to perform a question-answering task according to the question content and give a question-answering result; The target question and answer result conversion module is used to convert the target question and answer result based on the target question content into voice and return it to the target terminal device.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the smart phone outbound calling method for target position recruitment according to any one of claims 1-7.

10. A computer-readable medium, characterized in that The computer-readable medium stores computer instructions, and the computer instructions are used to enable a processor to implement the intelligent telephone outbound calling method for target position recruitment according to any one of claims 1 to 7 when executed.

Images