CN115955451A - Online session information safety monitoring system - Google Patents

Online session information safety monitoring system Download PDF

Info

Publication number
CN115955451A
CN115955451A CN202310223724.1A CN202310223724A CN115955451A CN 115955451 A CN115955451 A CN 115955451A CN 202310223724 A CN202310223724 A CN 202310223724A CN 115955451 A CN115955451 A CN 115955451A
Authority
CN
China
Prior art keywords
session
text
segment
management module
conversation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202310223724.1A
Other languages
Chinese (zh)
Other versions
CN115955451B (en
Inventor
杨吉伟
许柳飞
杜伟豪
陈健斌
梁伟锋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong Weixin Zhilian Technology Co ltd
Original Assignee
Guangdong Weixin Zhilian Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangdong Weixin Zhilian Technology Co ltd filed Critical Guangdong Weixin Zhilian Technology Co ltd
Priority to CN202310223724.1A priority Critical patent/CN115955451B/en
Publication of CN115955451A publication Critical patent/CN115955451A/en
Application granted granted Critical
Publication of CN115955451B publication Critical patent/CN115955451B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Machine Translation (AREA)

Abstract

The invention discloses an online session information safety monitoring system, which comprises: the system comprises a session management module, a data extraction module, a monitoring management module and an alarm module; the session management module sends the session segment to the data extraction module after the session starts; the data extraction module converts each conversation segment into a text segment; the data extraction module acquires the service attribute of the session according to the previous text segments and sends the service attribute to the client for confirmation; the data extraction module acquires keywords of a subsequent text segment according to the keyword dictionary; the data extraction module sends each text segment and the keywords thereof to the monitoring management module; the monitoring management module judges whether the own party is a multiparty party according to the text sections; if not, the monitoring management module acquires the conclusive text of each text segment of the other party in the conversation, judges whether the conclusive text is correct or not, and if not, sends the conclusive text to the alarm module; and the alarm module updates the risk value according to the conclusive text and sends the risk value to the session management module.

Description

Online session information safety monitoring system
Technical Field
The invention relates to the technical field of session information monitoring, in particular to an online session information safety monitoring system.
Background
With the development of on-line communication technology, more and more business activities are communicated and confirmed on line, but certain information difference exists between two communicated parties due to possibly known industry knowledge; this poor information leads to unfairness in business activities; for the information lacking party, the information difference between the information lacking party and the information multiple party is difficult to complement in a short time, and the risk of judging the online conversation by the information lacking party is not facilitated.
Disclosure of Invention
The invention aims to provide an online conversation information safety monitoring system and system, which solve the technical problem that the risk of an online conversation process is inconvenient to evaluate in the prior art.
The invention provides an online conversation information safety monitoring system, which comprises:
the system comprises a session management module, a data extraction module, a monitoring management module and an alarm module;
the session management module is used for responding to a session monitoring request of a client; after a session of a client starts, acquiring each session segment of a session interface of the client in real time and marking identity information to form a session stream and sequentially sending the session stream to a data extraction module; when the session management module receives a session ending instruction sent by the client, the session management module stops acquiring a session segment;
the data extraction module converts each conversation segment into a text segment; the data extraction module acquires the service attribute of the session according to the previous text segments and sends the service attribute to the client for confirmation; the data extraction module responds to the confirmation information of the client and then acquires a keyword dictionary corresponding to the service attribute; the data extraction module acquires keywords of a subsequent text segment according to the keyword dictionary and extracts the keywords of the previous text segments; the data extraction module sends each text segment and the corresponding keywords to the monitoring management module to form a text stream;
the monitoring management module acquires a plurality of text segments in the text stream at the initial stage of the session; the monitoring management module divides a plurality of text segments at the initial stage of the conversation into own text segments and other text segments through the identity information of each text segment and judges whether the own party is a multiparty party or not;
if not, the monitoring management module acquires the conclusive text of each text segment of the other party in the conversation, if so, the monitoring management module judges whether the conclusive text is correct or not by combining the keyword search result, and if not, the conclusive text is sent to the alarm module;
after the alarm module receives incorrect conclusive text, the risk value of the current session is updated and sent to the session management module;
the session management module displays the risk value on a session interface of the client; and when the risk value reaches the upper limit, reminding the user to end the conversation and limiting the conversation interface to pop up a financial transaction link.
Furthermore, the monitoring management module also retrieves professional words in the keywords of each text segment of the other party after the initial session through a distributed search engine; and generating a professional word analysis link according to the keyword retrieval result, and displaying the professional word analysis link on a session interface of the client so that the user can click and inquire.
Further, the data extraction module obtains the service attribute of the session according to the previous text segments, including:
obtaining verbs and nouns in each text section, and constructing a verb set and a noun set;
obtaining the similar meaning words of each verb in the verb set to expand the verb set; obtaining a near-meaning word of each noun in the noun set to expand the noun set;
constructing a retrieval formula according to the expanded verb set and the noun set; obtaining the service attribute of the session through a retrieval formula; the business attributes include consultation, business negotiation, contract making, and professionalism involved in the session.
Further, the monitoring management module judges whether the own party is the information multiparty, including:
acquiring the quantity and the types of the professional keywords in all own text sections in the initial session; the professional keywords comprise keywords related to the profession corresponding to the conversation;
acquiring the number and the types of the professional keywords in the text sections of all other parties in the initial session;
acquiring own professional value and other professional values according to the number and the types of the professional keywords;
and when the professional value of the own party is greater than that of the other party, judging that the own party is a plurality of information parties, wherein the plurality of information parties comprise main exporters of professional knowledge in the session process.
Further, the monitoring management module acquires conclusive texts of each other text segment in the conversation, including;
splitting each other-party text segment into a plurality of text sentences;
comparing each text sentence with a plurality of conclusion sentences one by one;
when a text sentence matches one of several conclusion sentences, then the text sentence is conclusive text.
Further, the alarm module updates the risk value of the current session, including:
obtaining corresponding positive words according to all key words and corresponding conclusion sentence patterns in the current conclusion text
An exact standard conclusion text;
obtaining conclusion modification adverbs in the current wrong conclusion text according to the correct standard conclusion text;
each conclusion modification adverb is provided with a corresponding risk value; and inputting the corresponding risk value into the risk value of the current conversation each time the conclusion modifier adverb appears.
Further, the background color of the conversation interface comprises green, yellow and red;
when the risk value of the session is in a first risk interval; the background color of the conversation interface is green;
when the risk value of the session is in the second risk interval; the background color of the conversation interface is yellow,
when the risk value of the session is in a third risk interval; the background color of the conversation interface is red,
the background color of the conversation interface of each risk interval is deepened along with the increase of the risk value.
Further, before determining whether the own party is the information multiple party, the method further comprises:
judging that the number of the text segments generated in the initial stage of the session is not less than a set value;
when the number of the text segments generated in the initial session stage is less than a set value, the monitoring management module acquires a plurality of dialect inquiry sentences and sends the plurality of dialect inquiry sentences to the session management module according to the own text segments and the dialect inquiry database; and the session management module sends the plurality of the session inquiry sentences to the client and displays the plurality of the session inquiry sentences on a session interface of the client for the selection of the client.
Further, the method also comprises that when the client sends a session monitoring request, the client sends the service attribute of the session to the session management module; and the session management module sends the information of the service attribute to the data extraction module, and the data extraction module acquires a keyword dictionary related to the service attribute and acquires keywords in the text segment according to the keyword dictionary.
Further, the data extraction module converts each conversation segment into a text segment, including:
judging the file attribute of the session segment;
when the file attribute of the conversation segment is audio, converting voice data in the audio into text data and forming a text segment;
and when the file attribute of the session segment is a picture, extracting text data in the picture and forming a text segment.
Further, still include: when the file attribute of the session segment is video:
screenshot is carried out on the video corresponding to the session section according to a preset period;
comparing the backgrounds of the screenshots; the method comprises the steps that a plurality of screenshots of the same background are reserved, wherein the screenshots with the highest definition are kept, and character recognition is carried out on the screenshots to obtain corresponding screenshot text data;
carrying out voice recognition on background audio in the video to acquire corresponding audio text data;
after the screenshot text data and the audio text data of the video are sorted, a text segment corresponding to the current conversation segment is obtained.
Compared with the prior art, the invention has the following beneficial effects:
the invention is connected with a client by setting a session management module, after the client sends a session monitoring request to the session management module, the session management module acquires each session segment of a session interface and sends the session segment to a data extraction module; the data extraction module acquires the corresponding business attribute of the session according to a plurality of previous session segments after the session starts so as to acquire a corresponding keyword dictionary, and acquires the keywords of the text segment according to the keyword dictionary so that the acquired keywords are more accurate; the monitoring management module judges whether the own party of the conversation is a multiparty information or not according to the text sections, and if the own party is not the multiparty information, the own party is an information deficient party; it is necessary to determine the correctness of the theoretical text in each other text segment. When more theoretical text errors exist in the text segment of the other party, the higher the risk of the session is proved to be, the alarm module evaluates the risk value, and the risk value is displayed on the client through the monitoring management module, so that the technical problem that the risk of the online session process is inconvenient to evaluate in the prior art is solved.
Drawings
Fig. 1 is a schematic block diagram of the present invention.
Detailed Description
As shown in fig. 1, an online session information security monitoring system includes:
the system comprises a session management module, a data extraction module, a monitoring management module and an alarm module;
the session management module is used for responding to a session monitoring request of a client; after a session of a client starts, acquiring each session segment of a session interface of the client in real time and marking identity information to form a session stream and sequentially sending the session stream to a data extraction module; when the session management module receives a session ending instruction sent by the client, the session management module stops acquiring a session segment;
the data extraction module converts each conversation segment into a text segment; the data extraction module acquires the service attribute of the session according to the previous text segments and sends the service attribute to the client for confirmation; the data extraction module responds to the confirmation information of the client and then acquires a keyword dictionary corresponding to the service attribute; the data extraction module acquires keywords of a subsequent text segment according to the keyword dictionary and extracts the keywords of the previous text segments; the data extraction module sends each text segment and the corresponding keywords to the monitoring management module to form a text stream;
the monitoring management module acquires a plurality of text segments in the text stream at the initial stage of the session; the monitoring management module divides a plurality of text segments at the initial stage of the conversation into own text segments and other text segments through the identity information of each text segment and judges whether the own party is a multiparty party or not;
if not, searching the professional words in the keywords of each text segment of the other party after the initial session through a distributed search engine; generating a professional word analysis link according to the keyword retrieval result, and displaying the professional word analysis link on a session interface of the client for a user to click and query;
the monitoring management module also acquires a conclusive text of each text segment of other parties in the conversation, if so, judges whether the conclusive text is correct or not by combining a keyword search result, and if not, sends the conclusive text to an alarm module;
after the alarm module receives incorrect conclusive text, the risk value of the current session is updated and sent to the session management module;
the session management module displays the risk value on a session interface of the client; and when the risk value reaches the upper limit, reminding the user to end the session and limiting the session interface to pop up a financial transaction link.
The specific implementation process of the embodiment includes:
in the actual online conversation process, information difference exists between two parties or one party and other parties of the conversation to a certain extent, the information difference causes that the information lacking party in the conversation process cannot know the conversation content, the information lacking party can be guided by the information difference, and the information lacking party cannot timely judge the safety and the correctness of the information in the conversation process. This is disadvantageous for fair disclosure of online sessions.
When a user needs to monitor and manage the current online session, sending a session monitoring request to a session management module, wherein the session monitoring request comprises a session ID of the opposite side; when new session information exists on a session interface of the client, the session management module collects the session information as session segments; the file attributes of the session segments of the user in the session interface comprise audio, pictures, texts and videos;
the data acquisition module converts each conversation segment into a text segment, and the method comprises the following steps:
judging the file attribute of the session segment;
when the file attribute of the conversation segment is audio, converting voice data in the audio into text data and forming a text segment;
and when the file attribute of the session segment is a picture, extracting text data in the picture and forming a text segment.
When the file attribute of the session segment is video:
screenshot is carried out on the video corresponding to the session section according to a preset period;
comparing the backgrounds of the screenshots; the method comprises the steps that a plurality of screenshots of the same background are reserved, wherein the screenshots with the highest definition are kept, and character recognition is carried out on the screenshots to obtain corresponding screenshot text data;
carrying out voice recognition on background audio in the video to acquire corresponding audio text data;
after the screenshot text data and the audio text data of the video are sorted, a text segment corresponding to the current conversation segment is obtained.
In the embodiment, the session management module is connected with the client, and after the client sends a session monitoring request to the session management module, the session management module acquires each session segment of a session interface and sends the session segment to the data extraction module; the data extraction module acquires the corresponding business attribute of the session according to a plurality of previous session segments after the session is started so as to acquire a corresponding keyword dictionary, and acquires the keywords of the text segment according to the keyword dictionary so that the acquired keywords are more accurate; the monitoring management module judges whether the own party of the conversation is a multiparty information or not according to the text sections, and if the own party is not the multiparty information, the own party is an information deficient party; the correctness of the theoretical text in each other text segment needs to be judged. When more theoretical text errors are generated in the text segment of the other party, the fact that the risk of the conversation is higher is proved, the alarm module evaluates the risk value, and the risk value is displayed on the client side through the monitoring management module, so that the technical problem that the risk of the online conversation process is inconvenient to evaluate in the prior art is solved.
In another embodiment of this embodiment, the obtaining, by the data extraction module, the service attribute of the current session according to the previous text segments includes:
obtaining verbs and nouns in each text segment, and constructing a verb set and a noun set;
obtaining the similar meaning words of each verb in the verb set to expand the verb set; obtaining a near-meaning word of each noun in the noun set to expand the noun set;
constructing a retrieval formula according to the expanded verb set and the noun set; acquiring the service attribute of the session through a retrieval formula; the business attributes include consultation, business negotiation, contract making, and professionalism involved in the session.
The specific implementation process of the embodiment includes:
after the conversation between the own party and the opposite party starts, the service attribute of the conversation can be obtained according to the previous conversation segments; in this embodiment, the service attribute of the session is obtained through the first ten session segments after the session starts, because for a session with a certain purpose, the session will be called from the beginning and quickly enter the session process of the formal topic, so that the service attribute of the session can be quickly judged through the first several session segments after the session starts.
In the embodiment, the nouns and verbs in the text segment are obtained, and the nouns and verbs are expanded to enlarge the retrieval range; the retrieval result is higher in accuracy, and then the retrieval result is sent to the session management module; the session management module sends the retrieval result to the client for confirmation and feeds the confirmation result back to the data acquisition module, and the data acquisition module acquires the corresponding keyword dictionary according to the feedback result.
In another embodiment of the invention, when sending a session monitoring request, a client sends the service attributes of the session to a session management module; and the session management module sends the information of the service attributes to the data extraction module, and the data extraction module acquires a keyword dictionary related to the service attributes and acquires keywords in the text segment according to the keyword dictionary. In this way, the data acquisition module can initially acquire the keywords of the text segment.
In another embodiment of the present invention, the determining whether the own party is the information party by the monitoring management module includes:
acquiring the quantity and the types of the professional keywords in all own text sections in the initial session; the professional keywords comprise keywords related to the profession corresponding to the conversation;
acquiring the quantity and the types of the professional keywords in all the text sections of other parties in the initial session;
acquiring own professional value and other professional values according to the number and the types of the professional keywords;
and when the professional value of the own party is greater than that of the other party, judging that the own party is a plurality of information parties, wherein the plurality of information parties comprise main exporters of professional knowledge in the session process.
The specific implementation process of the embodiment includes:
before judging whether the own party is the information multi-party, the method further comprises the following steps:
judging that the number of the text segments generated in the initial stage of the session is not less than a set value;
when the number of the text segments generated in the initial session stage is less than a set value, the monitoring management module acquires a plurality of session inquiry sentences and sends the plurality of session inquiry sentences to the session management module according to the own text segments and the session inquiry database; and the session management module sends the plurality of the session inquiry sentences to the client and displays the plurality of the session inquiry sentences on a session interface of the client for the selection of the client.
Obtaining enough text segments in the initial stage of the conversation is necessary for quickly and accurately judging whether the own party is a multi-party of information; when not enough text segments are generated within the beginning of a conversation,
and sending a plurality of dialect inquiry sentences to the client to remind the user to quickly enter the question and simultaneously obtain enough text segments.
The more the kind and number of professional words are present in the conversation segment, whether the own party or the other party; the specialization is stronger; the more specialized one tends to make information worse than the less specialized one. Especially in the case of own party being the party lacking information, the risk during the session is higher.
And if the own party is a plurality of parties, judging the correctness of the conclusive text of each text segment of the other party, marking the wrong conclusive text, feeding the marked conclusive text back to a session interface of the client through the session management module, and reminding a user of the client to correct the wrong conclusive text of the other party in time so as to avoid misunderstandings.
In this embodiment, obtaining the own-party professional value and the other-party professional value according to the number and the type of the professional keywords includes:
ZYH=α 1 *GSL+α 2 *GZL;
wherein ZYH is a professional value; GSL is the number of keywords; GZL is the category of the key words; alpha is alpha 1 And alpha 2 Is a scaling factor.
In another embodiment of the invention, the monitoring management module obtains the conclusive text of each other text segment in the conversation, including;
splitting each other text segment into a plurality of text sentences;
comparing each text sentence with a plurality of conclusion sentences one by one;
when a text sentence matches one of several conclusion sentences, then the text sentence is a conclusion text.
The specific implementation process of the embodiment includes:
judging whether the conclusive text sentence in the text segment is correct sentence by sentence; the conclusion sentence pattern includes: a is B; a must be B, A is absolutely B, and A is B.
In another embodiment of the present invention, the updating the risk value of the current session by the alarm module includes:
acquiring a corresponding correct standard conclusion text according to all key words and corresponding conclusion sentence patterns in the current conclusion text;
obtaining conclusion modification adverbs in the current wrong conclusion text according to the correct standard conclusion text;
each conclusion modification adverb is provided with a corresponding risk value; and inputting the corresponding risk value into the risk value of the current conversation each time the conclusion modifier adverb appears.
The specific implementation process of the embodiment includes:
incorrect conclusive text may increase confidence by modifying adverbs using some conclusion, such as absolute, impossible, supposed, approximate; the misleading strengths corresponding to different conclusion modification adverbs are different, namely the risk values are different; the higher the degree of certainty, the higher the risk value; and accumulating the risk value of the wrong conclusive text in the conversation process as the risk value of the conversation.
In another embodiment of the present invention, the background color of the session interface includes green, yellow and red;
when the risk value of the session is in a first risk interval; the background color of the conversation interface is green;
when the risk value of the session is in the second risk interval; the background color of the conversation interface is yellow,
when the risk value of the session is in a third risk interval; the background color of the conversation interface is red,
the background color of the conversation interface of each risk interval is deepened along with the increase of the risk value.
The present embodiment remarkably reminds the user of the risk condition of the current session by changing and deepening the background color of the session interface.
Although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the spirit and scope of the present invention.

Claims (10)

1. An online session information security monitoring system, characterized by: the system comprises:
the system comprises a session management module, a data extraction module, a monitoring management module and an alarm module;
the session management module is used for responding to a session monitoring request of a client; after a session of a client starts, acquiring each session segment of a session interface of the client in real time and marking identity information to form a session stream and sequentially sending the session stream to a data extraction module; when the session management module receives a session ending instruction sent by the client, the session management module stops acquiring a session segment;
the data extraction module converts each conversation segment into a text segment; the data extraction module acquires the service attribute of the session according to the previous text segments and sends the service attribute to the client for confirmation; the data extraction module responds to the confirmation information of the client and then acquires a keyword dictionary corresponding to the service attribute; the data extraction module acquires keywords of a subsequent text segment according to the keyword dictionary and extracts the keywords of the previous text segments; the data extraction module sends each text segment and the corresponding keywords to the monitoring management module to form a text stream;
the monitoring management module acquires a plurality of text segments in the text stream at the initial stage of the session; the monitoring management module divides a plurality of text segments at the initial stage of the conversation into own text segments and other text segments through the identity information of each text segment and judges whether the own party is a multiparty party or not;
if not, the monitoring management module acquires the conclusive text of each text segment of the other party in the conversation, judges whether the conclusive text is correct or not by combining the keyword search result, and if not, sends the conclusive text to the alarm module;
after the alarm module receives incorrect conclusive text, the risk value of the current session is updated and sent to the session management module;
the session management module displays the risk value on a session interface of the client; and when the risk value reaches the upper limit, reminding the user to end the session and limiting the session interface to pop up a financial transaction link.
2. The online session information security monitoring system of claim 1, wherein: the data extraction module acquires the service attribute of the session according to the previous text segments, and the method comprises the following steps:
obtaining verbs and nouns in each text segment, and constructing a verb set and a noun set;
obtaining the similar meaning words of each verb in the verb set to expand the verb set; obtaining the similar meaning word of each noun in the noun set to expand the noun set;
constructing a retrieval formula according to the expanded verb set and the noun set; acquiring the service attribute of the session through a retrieval formula; the business attributes include consultation, business negotiation, contract making, and professionalism involved in the session.
3. The online session information security monitoring system of claim 2, wherein: the monitoring management module judges whether the own party is the information multiparty or not, and comprises the following steps:
acquiring the quantity and the types of the professional keywords in all own text sections in the initial session; the professional keywords comprise keywords related to the profession corresponding to the conversation;
acquiring the number and the types of the professional keywords in the text sections of all other parties in the initial session;
acquiring own professional value and other professional values according to the number and the types of the professional keywords;
and when the professional value of the own party is greater than that of the other party, judging that the own party is a plurality of information parties, wherein the plurality of information parties comprise main exporters of professional knowledge in the session process.
4. The online session information security monitoring system of claim 1, wherein: the monitoring management module acquires conclusive texts of each other text segment in the conversation, including;
splitting each other text segment into a plurality of text sentences;
comparing each text sentence with a plurality of conclusion sentences one by one;
when a text sentence matches one of several conclusion sentences, then the text sentence is conclusive text.
5. The online session information security monitoring system of claim 4, wherein: the alarm module updates the risk value of the current session, and comprises the following steps:
obtaining corresponding correctness according to all key words and corresponding conclusion sentence patterns in the current concluding text
Standard conclusion text of (1);
obtaining conclusion modification adverbs in the current wrong conclusion text according to the correct standard conclusion text;
each conclusion modification adverb is provided with a corresponding risk value; and inputting the corresponding risk value into the risk value of the current conversation each time the conclusion modifier adverb appears.
6. The online session information security monitoring system of claim 1, wherein: the background colors of the conversation interface comprise green, yellow and red;
when the risk value of the session is in a first risk interval; the background color of the conversation interface is green;
when the risk value of the session is in the second risk interval; the background color of the conversation interface is yellow,
when the risk value of the session is in a third risk interval; the background color of the conversation interface is red,
the background color of the conversation interface of each risk interval is deepened along with the increase of the risk value.
7. The online session information security monitoring system of claim 3, wherein: before judging whether the own party is the information multi-party, the method further comprises the following steps:
judging that the number of the text segments generated in the initial stage of the session is not less than a set value;
when the number of the text segments generated in the initial session stage is less than a set value, the monitoring management module acquires a plurality of session inquiry sentences and sends the plurality of session inquiry sentences to the session management module according to the own text segments and the session inquiry database; and the session management module sends the plurality of the session inquiry sentences to the client and displays the plurality of the session inquiry sentences on a session interface of the client for the selection of the client.
8. The online session information security monitoring system of claim 1, wherein: when the client sends a session monitoring request, the client sends the service attributes of the session to the session management module; and the session management module sends the information of the service attributes to the data extraction module, and the data extraction module acquires a keyword dictionary related to the service attributes and acquires keywords in the text segment according to the keyword dictionary.
9. The online session information security monitoring system of claim 1, wherein:
the data extraction module converts each conversation segment into a text segment, and comprises the following steps:
judging the file attribute of the session segment;
when the file attribute of the conversation segment is audio, converting voice data in the audio into text data and forming a text segment;
and when the file attribute of the session segment is a picture, extracting text data in the picture and forming a text segment.
10. The online session information security monitoring system of claim 9, wherein: further comprising: when the file attribute of the session segment is video:
screenshot is carried out on the video corresponding to the session section according to a preset period;
comparing the backgrounds of the screenshots; the method comprises the steps that a plurality of screenshots of the same background are reserved, wherein the screenshots with the highest definition are kept, and character recognition is carried out on the screenshots to obtain corresponding screenshot text data;
carrying out voice recognition on background audio in the video to acquire corresponding audio text data;
after the screenshot text data and the audio text data of the video are sorted, a text segment corresponding to the current conversation segment is obtained.
CN202310223724.1A 2023-03-09 2023-03-09 Online session information security monitoring system Active CN115955451B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310223724.1A CN115955451B (en) 2023-03-09 2023-03-09 Online session information security monitoring system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310223724.1A CN115955451B (en) 2023-03-09 2023-03-09 Online session information security monitoring system

Publications (2)

Publication Number Publication Date
CN115955451A true CN115955451A (en) 2023-04-11
CN115955451B CN115955451B (en) 2023-07-14

Family

ID=85903313

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310223724.1A Active CN115955451B (en) 2023-03-09 2023-03-09 Online session information security monitoring system

Country Status (1)

Country Link
CN (1) CN115955451B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117456981A (en) * 2023-12-25 2024-01-26 北京秒信科技有限公司 Real-time voice wind control system based on RNN voice recognition

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9996531B1 (en) * 2016-03-29 2018-06-12 Facebook, Inc. Conversational understanding
WO2021042904A1 (en) * 2019-09-06 2021-03-11 平安国际智慧城市科技股份有限公司 Conversation intention recognition method, apparatus, computer device, and storage medium
CN113590802A (en) * 2021-09-27 2021-11-02 北京明略软件系统有限公司 Session content abnormity detection method and device, electronic equipment and storage medium
CN113822071A (en) * 2021-09-23 2021-12-21 秒针信息技术有限公司 Background information recommendation method and device, electronic equipment and medium
CN114861657A (en) * 2022-05-18 2022-08-05 北京金山数字娱乐科技有限公司 Conference key sentence extraction method and device

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9996531B1 (en) * 2016-03-29 2018-06-12 Facebook, Inc. Conversational understanding
WO2021042904A1 (en) * 2019-09-06 2021-03-11 平安国际智慧城市科技股份有限公司 Conversation intention recognition method, apparatus, computer device, and storage medium
CN113822071A (en) * 2021-09-23 2021-12-21 秒针信息技术有限公司 Background information recommendation method and device, electronic equipment and medium
CN113590802A (en) * 2021-09-27 2021-11-02 北京明略软件系统有限公司 Session content abnormity detection method and device, electronic equipment and storage medium
CN114861657A (en) * 2022-05-18 2022-08-05 北京金山数字娱乐科技有限公司 Conference key sentence extraction method and device

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117456981A (en) * 2023-12-25 2024-01-26 北京秒信科技有限公司 Real-time voice wind control system based on RNN voice recognition
CN117456981B (en) * 2023-12-25 2024-03-05 北京秒信科技有限公司 Real-time voice wind control system based on RNN voice recognition

Also Published As

Publication number Publication date
CN115955451B (en) 2023-07-14

Similar Documents

Publication Publication Date Title
Gleason et al. Making memes accessible
CN110493019B (en) Automatic generation method, device, equipment and storage medium of conference summary
US5740245A (en) Down-line transcription system for manipulating real-time testimony
US20170169822A1 (en) Dialog text summarization device and method
CN110459210A (en) Answering method, device, equipment and storage medium based on speech analysis
US20060078862A1 (en) Answer support system, answer support apparatus, and answer support program
US7797730B2 (en) Downline transcription system using automatic tracking and revenue collection
JP2017215931A (en) Conference support system, conference support device, conference support method, and program
US20070239446A1 (en) Down-line Transcription System Using Automatic Tracking And Revenue Collection
CN105912629A (en) Intelligent question and answer method and device
CN110895568B (en) Method and system for processing court trial records
CN105868179A (en) Intelligent asking-answering method and device
CN115955451B (en) Online session information security monitoring system
CN109101551B (en) Question-answer knowledge base construction method and device
US10789533B2 (en) Generating a consistently labeled training dataset by automatically generating and displaying a set of most similar previously-labeled texts and their previously assigned labels for each text that is being labeled for the training dataset
CN112699645B (en) Corpus labeling method, apparatus and device
CN111651554A (en) Insurance question-answer method and device based on natural language understanding and processing
CN110321414B (en) Artificial intelligence consultation service method and system based on deep learning
Love et al. Specifying challenges in transcribing covert recordings: Implications for forensic transcription
CN111061838A (en) Text feature keyword determination method and device and storage medium
CN113470625A (en) Voice conversation processing method, device, equipment and storage medium
JP7159576B2 (en) Information presentation device, information presentation system, information presentation method and program
CN111611387B (en) Civil case consultation method and device, computer equipment and storage medium
WO2023235580A1 (en) Video-based chapter generation for a communication session
CN114004599A (en) Material demand plan examination system based on artificial intelligence

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant