Conversational generation retrieval method and system based on large model document index perception

Abstract

The invention relates to the technical field of information retrieval, and provides a dialogue type generation retrieval method and system based on large model document index perception. The dialogue type generation retrieval method based on large model document index perception comprises the steps that according to user query input, a trained large language model is adopted, and an information sorting list related to the user query input is output; on the basis of a corpus, a cross attention layer is adopted, key information of a context is extracted, a proposition is obtained, and a document identifier is obtained; in the first-stage training, generating loss is introduced to generate information related to the current query, and a document identifier is generated; in the second-stage training, a ranking list of document identifiers retrieved by comprehensive loss optimization is introduced; and outputting a paragraph ranking list corresponding to the document identifier through a beam search decoding strategy. Through innovative document identifier design and training strategies, more effective context understanding and denoising are realized, and the dialogue type retrieval accuracy and user experience are improved.

Description

Technical Field

[0001] The present invention relates to the field of information retrieval technology, and in particular to a conversational generation and retrieval method and system based on large model document index perception. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] With the continuous development of artificial intelligence and natural language processing technology, conversational information retrieval systems have gradually become an important part of the modern information retrieval field. Conversational information retrieval aims to understand users' needs more accurately and provide relevant and accurate information through natural language interaction with users. Such systems can not only provide a more natural and intuitive search experience, but also dynamically adjust retrieval results based on users' contextual information and intent, thereby improving user satisfaction and retrieval efficiency. In traditional information retrieval systems, users usually obtain information by entering an independent query. However, in conversational information retrieval, users' queries are often contextual, which means that the model needs to extract key information from the conversation history to understand the user's current information needs. This poses unique challenges to traditional ad hoc retrieval methods, prompting the continuous development and innovation of conversational retrieval technology to better capture users' search intent and improve the accuracy of retrieval results. There are two main challenges facing conversational retrieval: First, user queries may contain pronoun references. For example, in a conversation, a user may use "it" or "the latter place" to refer to the previously mentioned "Dunkirk" and "Urk". Second, the intent of the current query may be obscured by irrelevant context. For example, in the last question about “Urk”, the retrieval model might be misled by the historical context related to “Dunkirk” and mistakenly rank documents about “Dunkirk” higher than those about “Urk”.

[0004] To address the above challenges, two main approaches have been developed: Contextual Query Rewrite (CQR) and Crash Data Retrieval (CDR). Conversational query rewriting approaches aim to train a rewriting model to transform the current query into an independent query based on the conversation context. However, since the rewriting process is independent of the retrieval process, these approaches have difficulties in end-to-end training and are limited by the query rewriting training data. Conversational dense retrieval approaches train a dual encoder model to embed all contexts and documents into a high-dimensional space to calculate relevance scores. Although some studies have designed complex training strategies to improve context representation, conversational dense retrieval approaches cannot achieve optimal denoising effects due to the limitations of dense retrieval. Specifically, dense retrieval uses fixed-length encodings to represent context, resulting in an embedding bottleneck problem. In addition, supervisory signals can only be propagated through these fixed-length encodings, limiting the model's learning ability. These inherent limitations are further exacerbated in conversational retrieval scenarios with complex and noisy contexts.

[0005] To overcome these limitations, generative retrieval is an emerging retrieval paradigm that utilizes a sequence-to-sequence framework to generate document identifiers (docids) for retrieval. Compared with dense retrieval, the main advantages of generative retrieval are two-fold. First, generative retrieval enables true end-to-end training, allowing supervisory signals to propagate directly to model parameters. This eliminates the embedding bottleneck problem, allowing the model to learn context denoising capabilities more effectively. Secondly, generative retrieval GR adopts an encoder-decoder architecture. In this framework, the cross-attention layer promotes the interaction between query tags and document tags, helping the decoder to better understand the input query, which is beneficial for context denoising. Although previous studies have demonstrated the advantages of generative retrieval in understanding complex user queries, previous methods only rely on document titles and do not consider the specific content of the document, resulting in poor performance on widely used conversational retrieval datasets. In summary, existing conversational retrieval methods have limitations in handling complex contexts and noise. Summary of the invention

[0006] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a conversational generative retrieval method and system based on large model document index perception. The present invention aims to improve the accuracy and robustness of conversational search in the presence of contextual noise through a generative retrieval framework based on large model document index perception, so as to better meet the information needs of users in the dynamic interaction process.

[0007] In order to achieve the above object, the present invention adopts the following technical solution: The first aspect of the present invention provides a conversational generation and retrieval method based on large model document index perception.

[0008] A conversational generative retrieval method based on large model document index perception, comprising: Based on the user query input, the trained large language model is used to output a ranked list of information related to the user query input; Among them, the process of training the large language model includes: based on the corpus, using the cross-attention layer to extract key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generation loss is introduced to generate information related to the current query and generate a document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; through the beam search decoding strategy, the paragraph ranking list corresponding to the document identifier is output.

[0009] Furthermore, the document identifier is a combination of the document title followed by a proposition.

[0010] Furthermore, during the decoding process of the large language model, the document title is used as a prefix to guide the encoding process.

[0011] Furthermore, the generation loss adopts maximum likelihood estimation loss.

[0012] Furthermore, the comprehensive loss is expressed by the following formula:

[0013]

[0014]

[0015] in, represents the comprehensive loss, represents the generation loss, represents the sorting loss, represents an arbitrary document identifier, Represents the first instance of the current training data hard negative document identifiers, is the margin hyperparameter, is the number of hard negative document identifiers, is a weight hyperparameter used to balance the two losses.

[0016] Furthermore, the beam search decoding strategy is used to output a paragraph ranking list corresponding to the document identifier; the method includes: constraining the decoding space through FM-index to ensure that the generated document identifier is valid, and ranking the paragraphs according to the generation probability to provide the user with the most relevant information to the user query input.

[0017] The second aspect of the present invention provides a conversational generation and retrieval system based on large model document index awareness.

[0018] A conversational generative retrieval system based on large-model document index perception, comprising: A query module is configured to: based on a user query input, use a trained large language model to output a ranked list of information related to the user query input; Among them, the process of training the large language model includes: based on the corpus, using the cross-attention layer to extract key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generation loss is introduced to generate information related to the current query and generate a document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; through the beam search decoding strategy, the paragraph ranking list corresponding to the document identifier is output.

[0019] A third aspect of the present invention provides a computer-readable storage medium.

[0020] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the conversational generation and retrieval method based on large model document index perception as described in the first aspect above.

[0021] A fourth aspect of the present invention provides a computer device.

[0022] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the conversational generation and retrieval method based on large model document index perception as described in the first aspect above are implemented.

[0023] A fifth aspect of the present invention provides a computer program product or a computer program.

[0024] The present invention provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the conversational generation and retrieval method based on large model document index perception as described in the first aspect above.

[0025] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes a conversational generative retrieval method and system based on large model document index perception. According to the user query input, the trained large language model is used to output a ranked list of information related to the user query input; wherein the process of large language model training includes: based on the corpus, the cross attention layer is used to extract the key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generative loss is introduced to generate information related to the current query and generate the document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of the retrieved document identifier; through the beam search decoding strategy, the paragraph ranking list corresponding to the document identifier is output. The present invention not only reduces the training burden by generating document identifiers, but also captures the fine-grained information in the paragraph, providing a new perspective for the application of generative retrieval in conversational search, and helping to locate and retrieve relevant information more accurately; through the two-stage training method, the ability of the large language model in context denoising and paragraph ranking has been comprehensively improved, and the most relevant information is provided to users under the constrained search decoding strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0027] Figure 1 It is a workflow diagram of the conversation-generated retrieval framework CGR4CD shown in the present invention. DETAILED DESCRIPTION

[0028] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0029] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0030] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0031] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the methods and systems according to various embodiments of the present disclosure. It should be noted that each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code may include one or more executable instructions for implementing the logical functions specified in each embodiment. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, or they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the flowchart and / or block diagram, and the combination of boxes in the flowchart and / or block diagram can be implemented using a dedicated hardware-based system that performs a specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0032] Embodiment 1 The present embodiment provides a conversational generation and retrieval method based on large model document index perception. The present embodiment uses the method applied to a server as an example for illustration. It is understandable that the method can also be applied to a terminal, and can also be applied to a terminal, a server, and a system, and is implemented through the interaction between the terminal and the server. The server can be an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network servers, cloud communications, middleware services, domain name services, security services CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited to this. The terminal and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in this application. In the present embodiment, the method includes the following steps: Based on the user query input, the trained large language model is used to output a ranked list of information related to the user query input; Among them, the process of training the large language model includes: based on the corpus, using the cross-attention layer to extract key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generation loss is introduced to generate information related to the current query and generate a document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; through the beam search decoding strategy, the paragraph ranking list corresponding to the document identifier is output.

[0033] The present invention designs the document identifier as a combination of a title and a proposition. Then, the large language model (LLM) is prompted to summarize the paragraph into multiple propositions to obtain the document identifier. During training, a two-stage training strategy is adopted to train the model step by step. For inference, a beam search is used to generate a ranked list, and FM-index is applied to constrain the decoding space to ensure that the generated document identifier is valid, such as Figure 1 As shown: The present embodiment is described in detail below: In some embodiments, the method is based on a corpus, adopts a cross attention layer, extracts key information of the context, obtains a proposition, and obtains a document identifier; the method comprises: The existing document identifier design in generative retrieval mainly faces two problems: it is difficult to expand to a large-scale corpus, and it is impossible to capture fine-grained information in a paragraph. To solve these problems, the present invention proposes to use title plus proposition as a document identifier. Propositions are defined as independent atomic expressions that convey key information in a paragraph. The inventors found that using propositions as document identifiers directly in generative retrieval can lead to unstable performance. This is because beam search can only explore a limited number of next tokens at each step, and in the initial decoding step, the real paragraph may be pruned incorrectly. To solve this problem, the title is used as a prefix to guide the decoding process, realizing retrieval from coarse-grained (title) to fine-grained (proposition). This greatly reduces the risk of incorrectly pruning real paragraphs.

[0034] In order to obtain propositions from paragraphs, a prompt word is designed to guide the large language model to summarize the paragraph into propositions, ensuring that each proposition is context-independent. For training data, the large language model is further instructed to select several propositions that can be used to answer the current question from real paragraphs as training targets.

[0035] Among them, in actual application, the prompt words in the above paragraph can be: "Your task is to summarize the content of the document into multiple clear and concise propositions, and ensure that these propositions are interpretable out of context."

[0036] In some embodiments, in the first stage of training, a generation loss is introduced to generate information related to the current query and generate a document identifier; the method includes: The present invention adopts a two-stage training strategy to train the model in a progressive manner. In the first stage, like other generative retrieval methods, an autoregressive generative model is trained using the standard maximum likelihood estimation (MLE) loss, with the input being the conversation context and the target being the document identifier. The goal of this stage is to train the model to recognize information related to the current query and generate the corresponding document identifier. For each training instance, the maximum likelihood estimation loss can be expressed as:

[0037] in, represents the maximum likelihood estimation loss (generation loss), is the number of real document identifiers, It is The length of the document identifier, Indicates document identifier, It means the of the document identifiers Tokens, Indicates the generation of retrieval model parameters The probability of generating Represents the conversation context.

[0038] In some embodiments, in the second stage training, a comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; the method includes: If the model is trained only with the maximum likelihood estimation loss, it tends to generate short, low-information propositions, such as "Dunkirk is a movie". To address this bias, a ranking loss is introduced in the second stage of training. Specifically, for each query of the training data instance, there will be a corresponding positive document identifier In order to construct the training data for the second stage, it is necessary to find a difficult negative sample for each training data instance. , by sampling document identifiers with the same title from all document identifiers in the corpus as this hard negative sample. In each training step, the generation probability of the document identifier is calculated , and train the model to assign higher probabilities to positive document identifiers. In addition, in order to make the training process more stable, the generation loss of the first stage is reintroduced. The loss of the second stage can be expressed as:

[0039]

[0040]

[0041] in, represents an arbitrary document identifier, Represents the first instance of the current training data hard negative document identifiers, represents the sorting loss, is the margin hyperparameter, is the number of hard negative document identifiers, represents the comprehensive loss, is a weight hyperparameter used to balance the two losses.

[0042] In some embodiments, the method of outputting a paragraph ranking list corresponding to a document identifier by a beam search decoding strategy comprises: In the inference phase, a constrained beam search decoding algorithm is used to generate a ranked list. In order to ensure that the generated document identifiers are valid, a data structure is required to store all propositions and constrain the decoding space. This is achieved using FM-index. FM-index supports returning a list of candidate tokens that satisfy the input prefix from any position. Therefore, the title and proposition of the paragraph are concatenated into the format of "document title@proposition1|document title@proposition2...", and the decoding is forced to start from the title. Finally, the paragraphs are ranked according to the probability of the generated document identifiers. The score of a paragraph can be calculated as:

[0043] in, Indicates The score of the paragraph, Indicates The set of document identifiers for all paragraphs.

[0044] This paper proposes a conversational generative retrieval method based on large-model document index perception, aiming to overcome the limitations of existing technologies and significantly improve the performance of conversational search systems in complex conversational scenarios through innovative model architecture and training strategies. First, a novel conversational generative retrieval framework (CGR4CD) is proposed: this framework uses a sequence-to-sequence generative retrieval model, and effectively captures key information in the context during the decoding process through a cross-attention layer, significantly improving the model's resistance to noise and irrelevant context, thereby more accurately understanding user intent and retrieving highly relevant information. At the same time, the powerful understanding and perception ability of the large model for text is used to build a fine-grained document index; a proposition-based document identifier (docids) is designed: compared with traditional document identifiers, the document identifier proposed in this invention not only reduces the training burden, but also captures the fine-grained information in the paragraph, providing a new perspective for the application of generative retrieval in conversational search, which helps to locate and retrieve relevant information more accurately; a two-stage training strategy is adopted: the strategy first trains the model through generation loss, so that it can identify information related to the current query from the context and generate corresponding document identifiers; then introduces ranking loss to further enhance the model's ability to distinguish similar paragraphs and optimize the ranking of retrieval results. This step-by-step training method has comprehensively improved the model's capabilities in context denoising and paragraph ranking; a constrained search decoding strategy is adopted in the inference stage: the decoding space is constrained by FM-index to ensure that the generated document identifier is valid, and the paragraphs are ranked according to the generation probability to provide users with the most relevant information.

[0045] Embodiment 2 This embodiment provides a conversational generation and retrieval system based on large model document index perception. A conversational generative retrieval system based on large-model document index perception, comprising: A query module is configured to: based on a user query input, use a trained large language model to output a ranked list of information related to the user query input; Among them, the process of training the large language model includes: based on the corpus, using the cross-attention layer to extract key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generation loss is introduced to generate information related to the current query and generate a document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; through the beam search decoding strategy, the paragraph ranking list corresponding to the document identifier is output.

[0046] In some embodiments, the document identifier is a combination of a document title followed by a proposition.

[0047] In some embodiments, during the large language model decoding process, the document title is used as a prefix to guide the encoding process.

[0048] In some embodiments, the generation loss adopts maximum likelihood estimation loss.

[0049] In some embodiments, the comprehensive loss is expressed by the following formula:

[0050]

[0051]

[0052] in, represents the comprehensive loss, represents the generation loss, represents the sorting loss, represents an arbitrary document identifier, Represents the first instance of the current training data hard negative document identifiers, is the margin hyperparameter, is the number of hard negative document identifiers, is a weight hyperparameter used to balance the two losses.

[0053] In some embodiments, the beam search decoding strategy is used to output a paragraph ranking list corresponding to the document identifier; the method includes: using FM-index to constrain the decoding space to ensure that the generated document identifier is valid, and ranking the paragraphs according to the generation probability to provide the user with the most relevant information to the user's query input.

[0054] The present invention achieves more effective context understanding and denoising through innovative document identifier design and training strategies, thereby improving the accuracy of conversational retrieval and user experience.

[0055] Embodiment 3 This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the conversational generation and retrieval method based on large model document index perception as described in the first embodiment above are implemented.

[0056] Embodiment 4 This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the conversational generation and retrieval method based on large model document index perception as described in the first embodiment above are implemented.

[0057] Embodiment 5 This embodiment provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps of the conversational generation and retrieval method based on large model document index perception described in the first embodiment.

[0058] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0059] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0060] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0061] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0062] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0063] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A conversational generative retrieval method based on large-model document index perception, characterized in that: include: Based on the user query input, the trained large language model is used to output a ranked list of information related to the user query input; Among them, the process of training the large language model includes: based on the corpus, using the cross-attention layer to extract key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generation loss is introduced to generate information related to the current query and generate a document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; through the beam search decoding strategy, the paragraph ranking column corresponding to the document identifier is output.

2. The conversational generation and retrieval method based on large model document index perception according to claim 1 is characterized in that: The document identifier is a combination of the document title followed by a proposition.

3. The conversational generation and retrieval method based on large model document index perception according to claim 1 is characterized in that: In the large language model decoding process, the document title is used as a prefix to guide the encoding process.

4. The conversational generation and retrieval method based on large model document index perception according to claim 1 is characterized in that: The generation loss adopts the maximum likelihood estimation loss.

5. The conversational generation and retrieval method based on large model document index perception according to claim 1 is characterized in that: The comprehensive loss is expressed by the following formula: in, represents the comprehensive loss, represents the generation loss, represents the sorting loss, represents an arbitrary document identifier, Represents the current training data instance hard negative document identifiers, is the margin hyperparameter, is the number of hard negative document identifiers, is a weight hyperparameter used to balance the two losses.

6. The conversational generation and retrieval method based on large model document index perception according to claim 1 is characterized in that: The beam search decoding strategy is used to output a paragraph ranking list corresponding to the document identifier; the method includes: constraining the decoding space through FM-index to ensure that the generated document identifier is valid, and ranking the paragraphs according to the generation probability to provide the user with the most relevant information to the user query input.

7. A conversational generative retrieval system based on large-model document index perception, characterized in that: include: A query module is configured to: based on a user query input, use a trained large language model to output a ranked list of information related to the user query input; Among them, the process of training the large language model includes: based on the corpus, using the cross-attention layer to extract key information of the context, obtain the proposition, and obtain the document identifier; in the first stage of training, the generation loss is introduced to generate information related to the current query and generate a document identifier; in the second stage of training, the comprehensive loss is introduced to optimize the ranking list of retrieved document identifiers; through the beam search decoding strategy, the paragraph ranking list corresponding to the document identifier is output.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the conversational generation and retrieval method based on large model document index perception as described in any one of claims 1-6 are implemented.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the steps in the conversational generation and retrieval method based on large model document index perception as described in any one of claims 1-6.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps in the conversational generation and retrieval method based on large model document index perception as described in any one of claims 1-6.

Images