CN-121998604-A - Mail reply content generation method based on knowledge base and computer program product

Abstract

The invention relates to the technical field of artificial intelligence, in particular to a mail reply content generation method and a computer program product based on a knowledge base, comprising the steps of obtaining mail data to be replied; determining a first threshold and a summary degree range according to the mail data to be replied, wherein the first threshold is a similarity threshold of the data in a user knowledge base and the mail data to be replied, the summary degree is a ratio of the number of words after summarization to the number of words before summarization, calculating the similarity of the data in the user knowledge base and the mail data to be replied to obtain matching similarity, screening the data with the matching similarity exceeding the first threshold in the user knowledge base to obtain target data, and calling AI to generate reply mails conforming to the summary degree range according to the target data and the mail data to be replied. The proposal of the invention can automatically and effectively reply the mail and gives consideration to the information control accuracy.

Inventors

• Request for anonymity
• LU ZHENG

Assignees

• 安徽声云智能科技有限公司

Dates

Publication Date

20260508

Application Date

20260128

Claims (10)

1. 1. A knowledge base-based mail reply content generation method, the method comprising: Acquiring mail data to be replied; Determining a first threshold and a summary range according to the mail data to be replied, wherein the first threshold is a similarity threshold of the data in a user knowledge base and the mail data to be replied, and the summary is a ratio of the number of words after summarization to the number of words before summarization; calculating the similarity between the data in the user knowledge base and the mail data to be replied to obtain matching similarity; Screening data with matching similarity exceeding the first threshold value in the user knowledge base to obtain target data; And invoking AI (advanced technology interface) to generate a reply mail conforming to the summary range according to the target data and the mail data to be replied.
2. 2. The method of claim 1, wherein the first threshold and the summary range are obtained by: the method comprises the steps of obtaining background data, wherein the background data comprises relationship data of a compound to be returned and a user side, identity data of the compound to be returned and identity data of the user side; And invoking AI to determine the first threshold and the summarization range according to the mail data to be replied and the background data.
3. 3. The method of claim 2, wherein the user knowledge base comprises an enterprise knowledge base and/or a personal knowledge base, the method further comprising: and sending the first threshold value and the summary range determined by the AI to the user, and carrying out subsequent steps after the user modifies or confirms the first threshold value and the summary range.
4. 4. The method of claim 1, wherein the first threshold and the summary range are obtained by: And determining the first threshold and the summarization range according to the mail data to be replied by the user.
5. 5. The method according to claim 4, wherein the method further comprises: Presetting at least two privacy levels, wherein each privacy level corresponds to a set first threshold value and a summary degree range; The user determines the first threshold and the summary range by selecting a privacy level.
6. 6. The method of claim 5, wherein the privacy levels comprise a highest privacy level and a lowest privacy level; at the highest privacy level, the summary range is 0, and the reply mail does not contain content related to the target data; At the lowest privacy level, the summary range is 100%, and the reply mail contains all contents of the target data.
7. 7. The method according to claim 1, wherein the method further comprises: Acquiring historical sent mails of a user; and the AI is required to adjust the language style of the reply mail to be consistent with the historical sent mail according to the historical sent mail.
8. 8. The method according to claim 1, wherein the method further comprises: and calculating the similarity between the data in the user knowledge base and the mail data to be replied by using a similarity algorithm, wherein the value range of the summary degree range is 0-100%, and the value range of the first threshold value is 0-100%.
9. 9. The method according to claim 4, wherein the method further comprises: And setting a first control bar and a second control bar on the user interaction interface, wherein the first control bar is used for setting the first threshold value by a user, and the second control bar is used for setting the summarization range by the user.
10. 10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any one of claims 1 to 9.

Description

Mail reply content generation method based on knowledge base and computer program product Technical Field The present disclosure relates to the field of artificial intelligence, and in particular, to a method and a computer program product for generating mail reply content based on a knowledge base. Background This section is intended to provide a background or context for the disclosed embodiments. The description herein is not admitted to be prior art by inclusion in this section. With the rapid growth of email traffic, efficient handling of email has become a key to improving work efficiency. The main mail reply modes at present comprise manual word-by-word reply, low efficiency and high dependence on manual work, fixed template reply, lack of individuation and pertinence, and an automatic reply system based on rules, which has poor flexibility and cannot adapt to diversified mail contents. In recent years, artificial intelligence (AI for short, english full name ARTIFICIAL INTELLIGENCE) technology has been applied to mail automatic reply. For example, some intelligent reply systems generate reply suggestions through natural language processing techniques, and platforms such as the ali cloud also provide similar automated services. However, these prior arts have significant drawbacks in that firstly, the generated reply contents are highly generalized and difficult to reply effectively for specific problems, and secondly, the system cannot control the information disclosure degree intelligently, which may cause excessive disclosure of confidential information or insufficient disclosure of confidential information in business transactions to affect communication efficiency. Disclosure of Invention In view of the foregoing, it is desirable to provide a knowledge base-based mail reply content generation method and a computer program product that can automatically and effectively reply to a mail, while taking account of the accuracy of information control. In a first aspect, the present disclosure provides a method for generating mail reply content based on a knowledge base. The method comprises the following steps: Acquiring mail data to be replied; Determining a first threshold and a summary range according to the mail data to be replied, wherein the first threshold is a similarity threshold of the data in a user knowledge base and the mail data to be replied, and the summary is a ratio of the number of words after summarization to the number of words before summarization; calculating the similarity between the data in the user knowledge base and the mail data to be replied to obtain matching similarity; Screening data with matching similarity exceeding the first threshold value in the user knowledge base to obtain target data; And invoking AI (advanced technology interface) to generate a reply mail conforming to the summary range according to the target data and the mail data to be replied. Optionally, the first threshold and the summary range are obtained by: the method comprises the steps of obtaining background data, wherein the background data comprises relationship data of a compound to be returned and a user side, identity data of the compound to be returned and identity data of the user side; And invoking AI to determine the first threshold and the summarization range according to the mail data to be replied and the background data. Optionally, the user knowledge base comprises an enterprise knowledge base and/or a personal knowledge base, and the method further comprises: and sending the first threshold value and the summary range determined by the AI to the user, and carrying out subsequent steps after the user modifies or confirms the first threshold value and the summary range. Optionally, the first threshold and the summary range are obtained by: And determining the first threshold and the summarization range according to the mail data to be replied by the user. Optionally, the method further comprises: Presetting at least two privacy levels, wherein each privacy level corresponds to a set first threshold value and a summary degree range; The user determines the first threshold and the summary range by selecting a privacy level. Optionally, the privacy level includes a highest privacy level and a lowest privacy level; at the highest privacy level, the summary range is 0, and the reply mail does not contain content related to the target data; At the lowest privacy level, the summary range is 100%, and the reply mail contains all contents of the target data. Optionally, the method further comprises: Acquiring historical sent mails of a user; and the AI is required to adjust the language style of the reply mail to be consistent with the historical sent mail according to the historical sent mail. Optionally, the method further comprises: and calculating the similarity between the data in the user knowledge base and the mail data to be replied by using a similarity algorithm, wherein the value rang