CN107784033B - Method and device for recommending based on session - Google Patents
Method and device for recommending based on session Download PDFInfo
- Publication number
- CN107784033B CN107784033B CN201610798166.1A CN201610798166A CN107784033B CN 107784033 B CN107784033 B CN 107784033B CN 201610798166 A CN201610798166 A CN 201610798166A CN 107784033 B CN107784033 B CN 107784033B
- Authority
- CN
- China
- Prior art keywords
- current
- knowledge
- answer
- client
- session
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
Abstract
The invention aims to provide a method and a device for real-time recommendation based on a session, which are characterized in that on the basis of the current session between a current user and a current client, according to the attribute information of the current client, a recommendation answer corresponding to the current question in the current session is obtained by matching in a knowledge base; providing the recommended answer to the current user; wherein the structure of knowledge in the knowledge base comprises < customer attributes, questions, answers >; compared with the prior art, the method and the system effectively utilize the prior knowledge, are convenient for the current user to answer the inquiry of the current client, and improve the use experience of the user.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a technology for recommending based on a session.
Background
For a department or company that needs a lot of voice calls, it accumulates a lot of voice materials including much experience and knowledge. But due to lack of effective management, newly entered employees are difficult to benefit from. One major problem is that manual speech translation requires a lot of manpower and material resources, and it is not easy for new employees to be skilled in the extracted knowledge.
The existing scheme mainly depends on manual knowledge extraction and user self memory and application, however, the existing method has the following disadvantages: 1) the manual extraction is difficult to ensure the comprehensiveness of knowledge, and completely depends on personal experience, so that good knowledge can be ignored, or bad knowledge can be extracted, and the optimal answer is difficult to ensure; 2) the memory of knowledge requires time, the workload of a user is increased, and correct knowledge cannot be memorized at a proper time; 3) knowledge cannot be optimized, good knowledge cannot get more exposures, and bad knowledge does not reduce the number of exposures.
Therefore, how to make answer recommendations based on sessions becomes one of the problems that those skilled in the art need to solve.
Disclosure of Invention
The invention aims to provide a method and a device for recommending based on a session.
According to an aspect of the present invention, there is provided a method for making a recommendation based on a session, wherein the method comprises the steps of:
based on a current session between a current user and a current client thereof, matching and obtaining a recommended answer corresponding to a current question in the current session from a knowledge base according to the attribute information of the current client;
providing the recommended answer to the current user;
wherein the structure of knowledge in the knowledge base comprises < customer attributes, questions, answers >.
Preferably, the knowledge base is built or updated using the following process:
acquiring historical session records between each user and the client of the user;
extracting knowledge from the historical conversation records to obtain corresponding knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers;
and establishing or updating the knowledge base according to the knowledge.
Preferably, the step of extracting knowledge from the historical session record to obtain corresponding knowledge includes:
converting the historical conversation record into corresponding conversation text information;
and extracting knowledge from the session text information to obtain the knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers.
More preferably, the step of extracting knowledge from the session text information to obtain corresponding knowledge includes:
abstracting the session text information into a binary knowledge structure, wherein the binary knowledge structure comprises < user or client, say >;
recombining the binary knowledge structure into a ternary knowledge structure, wherein the ternary knowledge structure comprises < client, question, answer >;
and clustering each element in the ternary knowledge structure to obtain the knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers.
Preferably, the method further comprises:
determining the priority of the recommended answer according to the occurrence times of the recommended answer and the current question corresponding to the recommended answer in the knowledge base;
wherein the step of providing the recommended answer to the current user comprises:
and providing the recommended answer to the current user according to the priority.
Preferably, the method further comprises:
and adjusting the priority of the recommended answer according to the actual answer of the current user to the current question in the current session.
Preferably, the method further comprises:
and modifying the attribute information of the current client in real time according to the question and answer of the current user and the current client in the current session.
Preferably, the method further comprises:
and acquiring the current session between the current user and the current client as a historical session record so as to extract corresponding knowledge.
According to another aspect of the present invention, there is also provided a recommendation apparatus for making a recommendation based on a session, wherein the recommendation apparatus includes:
the device is used for matching and obtaining a recommended answer corresponding to a current question in the current conversation from a knowledge base according to the attribute information of the current client based on the current conversation between the current user and the current client;
means for providing the recommended answer to the current user;
wherein the structure of knowledge in the knowledge base comprises < customer attributes, questions, answers >.
Preferably, the recommendation apparatus further comprises:
means for obtaining historical session records between each user and their clients;
a device for extracting knowledge from the historical conversation record to obtain corresponding knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers;
means for building or updating the knowledge base based on the knowledge;
preferably, the device for extracting knowledge from the historical session record to obtain corresponding knowledge includes:
means for converting the historical conversation record into corresponding conversation text information;
and a unit for extracting knowledge from the session text information to obtain the knowledge, wherein the structure of the knowledge comprises < customer attribute, question, answer >.
More preferably, the unit for extracting knowledge from the session text information and obtaining the knowledge is configured to:
abstracting the session text information into a binary knowledge structure, wherein the binary knowledge structure comprises < user or client, say >;
recombining the binary knowledge structure into a ternary knowledge structure, wherein the ternary knowledge structure comprises < client, question, answer >;
and clustering each element in the ternary knowledge structure to obtain the knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers.
Preferably, the recommendation apparatus further comprises:
means for determining a priority of the recommended answer based on the number of occurrences of the recommended answer and its corresponding current question in the knowledge base;
wherein the means for providing the recommended answer to the current user is to:
and providing the recommended answer to the current user according to the priority.
Preferably, the recommendation apparatus further comprises:
means for adjusting a priority of the recommended answer according to an actual answer of the current user to the current question in the current session.
Preferably, the recommendation apparatus further comprises:
and modifying the attribute information of the current client in real time according to the question and answer of the current user and the current client in the current session.
Preferably, the recommendation apparatus further comprises:
and the device is used for acquiring the current session between the current user and the current client as a historical session record so as to extract corresponding knowledge.
According to still another aspect of the present invention, there is also provided a computer apparatus including:
one or more processors;
a memory for storing one or more computer programs;
the one or more computer programs, when executed by the one or more processors, cause the one or more processors to implement the methods as described above.
Compared with the prior art, the method and the device have the advantages that based on the current session between the current user and the current client, the recommended answer corresponding to the current question in the current session is obtained from the knowledge base in a matching mode according to the attribute information of the current client, and is provided for the current user, wherein the structure of the knowledge in the knowledge base comprises the attributes, the questions and the answers of the client, the existing knowledge is effectively utilized, the current user can conveniently answer the inquiry of the current client, and the use experience of the user is improved.
Furthermore, the invention extracts knowledge from the historical conversation records between each user and the client thereof, and stores the knowledge in a mode of < client attribute, question and answer >, so that when a subsequent user communicates with the client, the recommended answer is matched from the historical conversation records to the user for answering according to the attribute information of the client, thereby realizing the management of the knowledge on one hand, and carrying out the knowledge recommendation in real time according to the context on the other hand, improving the system efficiency and improving the use experience of the user.
Furthermore, the invention can also determine the priority of the recommended answer and the occurrence frequency of the current question corresponding to the recommended answer in the knowledge base, and provide the recommended answer to the current user according to the priority, thereby further improving the use experience of the user. Furthermore, the invention can adjust the priority of the recommended answer according to whether the actual answer of the current user is consistent with the recommended answer or not, effectively utilizes the actual feedback of the user and further improves the use experience of the user.
Furthermore, the attribute information of the client is modified in real time according to the question and answer of the current client, so that the subsequent matching operation is more accurate, and the use experience of the user is further improved.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments made with reference to the following drawings:
FIG. 1 illustrates a block diagram of an apparatus for session-based recommendation in accordance with an aspect of the present invention;
FIG. 2 is a block diagram of an apparatus for session-based recommendation according to an embodiment of the present invention;
FIG. 3 illustrates a flow diagram of a method for session-based recommendation in accordance with another aspect of the subject invention;
FIG. 4 shows a flow diagram of a method for session-based recommendation, according to another embodiment of the present invention.
The same or similar reference numbers in the drawings identify the same or similar elements.
Detailed Description
Before discussing exemplary embodiments in more detail, it should be noted that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although a flowchart may describe the operations as a sequential process, many of the operations can be performed in parallel, concurrently, or simultaneously. In addition, the order of the operations may be re-arranged. The process may be terminated when its operations are completed, but may have additional steps not included in the figure. The processes may correspond to methods, functions, procedures, subroutines, and the like.
The term "computer device" or "computer" in this context refers to an intelligent electronic device that can execute predetermined processes such as numerical calculation and/or logic calculation by running predetermined programs or instructions, and may include a processor and a memory, wherein the processor executes a pre-stored instruction stored in the memory to execute the predetermined processes, or the predetermined processes are executed by hardware such as ASIC, FPGA, DSP, or a combination thereof. Computer devices include, but are not limited to, servers, personal computers, laptops, tablets, and the like.
The computer equipment comprises user equipment and network equipment. Wherein the user equipment includes but is not limited to personal computers, notebook computers, tablet computers, and the like; the network device includes, but is not limited to, a single network server, a server group consisting of a plurality of network servers, or a Cloud Computing (Cloud Computing) based Cloud consisting of a large number of computers or network servers, wherein Cloud Computing is one of distributed Computing, a super virtual computer consisting of a collection of loosely coupled computers. Wherein the computer device can be operated alone to implement the invention, or can be accessed to a network and implement the invention through interoperation with other computer devices in the network. The network in which the computer device is located includes, but is not limited to, the internet, a wide area network, a metropolitan area network, a local area network, a VPN network, and the like.
It should be noted that the user equipment, the network device, the network, etc. are only examples, and other existing or future computer devices or networks may also be included in the scope of the present invention, and are included by reference.
The methods discussed below, some of which are illustrated by flow diagrams, may be implemented by hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When implemented in software, firmware, middleware or microcode, the program code or code segments to perform the necessary tasks may be stored in a machine or computer readable medium such as a storage medium. The processor(s) may perform the necessary tasks.
Specific structural and functional details disclosed herein are merely representative and are provided for purposes of describing example embodiments of the present invention. The present invention may, however, be embodied in many alternate forms and should not be construed as limited to only the embodiments set forth herein.
It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element may be termed a second element, and, similarly, a second element may be termed a first element, without departing from the scope of example embodiments. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may be present. In contrast, when an element is referred to as being "directly connected" or "directly coupled" to another element, there are no intervening elements present. Other words used to describe the relationship between elements (e.g., "between" versus "directly between", "adjacent" versus "directly adjacent to", etc.) should be interpreted in a similar manner.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used herein, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It should also be noted that, in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed substantially concurrently, or the figures may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
The present invention is described in further detail below with reference to the attached drawing figures.
Fig. 1 illustrates a block diagram of an apparatus for session-based recommendation in accordance with an aspect of the present invention.
The recommendation device 1 is located in a computer device, for example, and the recommendation device 1 includes a device for matching and obtaining a recommendation answer corresponding to a current question in a current session from a knowledge base according to attribute information of a current client based on the current session between a current user and the current client, which is hereinafter referred to as matching device 101; and a device for providing the recommended answer to the current user, hereinafter referred to as a providing device 102, wherein the structure of the knowledge in the knowledge base includes < customer attribute, question, answer >.
The matching device 101 matches and obtains a recommended answer corresponding to a current question in the current session from the knowledge base according to attribute information of the current client based on the current session between the current user and the current client, wherein the structure of knowledge in the knowledge base comprises < client attribute, question, answer >.
Specifically, various types of knowledge are stored in the knowledge base, for example, the knowledge is stored in the knowledge base in a structure of < client attribute, question, answer >, and the knowledge is extracted from a large number of users according to the experience of the conversation with the clients thereof. For example, when a customer with a certain attribute asks a certain question, what the given answer is, the other users can refer to how to answer the customer's question conveniently by counting, sorting and storing the questions or answers in the knowledge base. The knowledge base can store various kinds of knowledge in a voice form or a character form, and can be located in the recommendation device 1 or in a third-party device connected with the recommendation device 1 through a network.
When a current user performs a current session with a current client thereof, for example, the current user is performing a voice call, the matching device 101 matches and obtains a recommended answer corresponding to a current question in the current session from the knowledge base according to the attribute information of the current client, for example, the matching device 101 first matches and obtains a client attribute corresponding to the attribute information and all knowledge corresponding to the client attribute in the knowledge base according to the attribute information of the current client, such as the attribute information of the industry, the region, the scale, and the like of the current client, and then matches and obtains a recommended answer corresponding to the question in the knowledge according to the current question asked by the current client in the current session.
Here, the attribute information of the current client is, for example, known to the current user, for example, known before the current user performs the current session, or may be attribute information analyzed or adjusted in real time by the recommendation apparatus 1.
Here, the matching device 101 may also convert the current question in the current session into text information through voice recognition, and then perform matching search in the knowledge base according to the text information.
It should be understood by those skilled in the art that the above-mentioned method for matching recommended answers is only an example, and other methods for matching recommended answers that are currently available or that may be later developed, such as being applicable to the present invention, are also included in the scope of the present invention and are also included herein by reference.
The providing device 102 provides the recommended answer to the current user.
Specifically, for the recommended answer obtained by matching by the matching device 101, the providing device 102 provides the recommended answer to the current user for the current user to select, for example, through a presentation manner of voice, text, or other convention. For example, when the current user uses a wearable device to perform a current session with the current client, the providing device 102 interacts with the wearable device, and provides the recommended answer matched by the matching device 101 to the current user via the wearable device. Preferably, for the current question in the current session, if the matching device 101 matches multiple recommended answers, the providing device 102 may provide the multiple recommended answers to the user randomly or in a certain order.
It should be understood by those skilled in the art that the above-mentioned manner for providing recommended answers is only an example, and other manners for providing recommended answers that are currently available or that may come out later, such as may be applicable to the present invention, are also included in the scope of the present invention and are herein incorporated by reference.
The recommendation device 1 matches and obtains a recommendation answer corresponding to a current question in a current session from a knowledge base according to attribute information of a current client based on the current session between the current user and the current client, and provides the recommendation answer to the current user, wherein the structure of knowledge in the knowledge base comprises < client attribute, question and answer >, so that the current knowledge is effectively utilized, the current user can conveniently answer the inquiry of the current client, and the use experience of the user is improved.
Fig. 2 is a schematic structural diagram of an apparatus for performing a recommendation based on a session according to an embodiment of the present invention.
The recommendation device 1 further comprises means for acquiring historical session records between each user and its client, hereinafter referred to as first acquisition means 203; a device for extracting knowledge from the historical session record to obtain corresponding knowledge, wherein the structure of the knowledge includes < customer attribute, question, answer >, hereinafter referred to as knowledge extraction device 204; and means for establishing or updating said knowledge base based on said knowledge, hereinafter referred to as updating means 205. The matching device 201 and the providing device 202 are the same or substantially the same as the corresponding devices shown in fig. 1, and therefore are not described herein again and are included herein by way of reference.
Wherein the first obtaining device 203 obtains the history session record between each user and the client thereof.
Specifically, when the user communicates with the client, the content of the communication may be recorded, for example, when the user makes a call with the client, the voice call content is recorded by recording through the voice call device used by the user, and the first obtaining device 203 obtains a plurality of historical conversation records between each user and the client regularly or in real time, for example, through interaction with each voice call device.
It should be understood by those skilled in the art that the above-mentioned manner of obtaining historical session records is only an example, and other existing or future manners of obtaining historical session records, such as those that may be applicable to the present invention, are also included within the scope of the present invention and are incorporated herein by reference.
The knowledge extraction device 204 performs knowledge extraction on the historical session records to obtain corresponding knowledge, wherein the structure of the knowledge includes < customer attribute, question, answer >.
Specifically, for the historical conversation record acquired by the first acquisition device 203, the knowledge extraction device 204 performs knowledge extraction on the historical conversation record, for example, first converts the historical conversation record in the voice form into a historical conversation record in the text form, and then extracts knowledge from the historical conversation record in the text form, where the structure of the knowledge includes < client attribute, question, answer >, and for example, the question or answer corresponding to the client is classified according to the client attribute corresponding to different clients, and then the question and answer belonging to one conversation are taken as one knowledge, so as to obtain the knowledge with the structure of < client attribute, question, answer >.
Preferably, the knowledge extraction means 204 comprises a conversion unit (not shown) and an extraction unit (not shown). The conversion unit converts the historical conversation record into corresponding conversation text information; and the extraction unit extracts the knowledge of the session text information to obtain the knowledge, wherein the structure of the knowledge comprises the client attribute, the question and the answer.
Specifically, for the history session record in the voice form acquired by the first acquisition device 203, the conversion unit converts the history session record in the voice form into corresponding session text information by using a voice recognition technology, a character conversion technology, or the like; further, the conversion unit may also first perform preprocessing on the historical conversation record acquired by the first acquiring device 203, such as determining the language family of the historical conversation record, determining the front and back nasal sounds of the historical conversation record, and the like, to obtain preprocessed information; and determining corresponding session text information according to the historical session record or further combining the user related information and the preprocessing information.
Subsequently, the extraction unit extracts the knowledge of the session text information converted by the conversion unit to obtain the knowledge, for example, classifying the questions or answers corresponding to the clients according to the client attributes corresponding to different clients, and then using the questions and answers belonging to one session as one knowledge, thereby obtaining the knowledge having the structure of < client attributes, questions, answers >.
More preferably, the extraction unit abstracts the session text information into a binary knowledge structure, wherein the binary knowledge structure includes < user or client, say >; recombining the binary knowledge structure into a ternary knowledge structure, wherein the ternary knowledge structure comprises < client, question, answer >; and clustering each element in the ternary knowledge structure to obtain the knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers.
Specifically, after the conversion unit converts the historical conversation record into the conversation text information, the extraction unit abstracts the conversation text information into a binary knowledge structure, for example, by means of classification, wherein the binary knowledge structure includes < user or client, say >, where the say may be a question or an answer, that is, the extraction unit classifies the conversation text information into < user, question >, < user, answer >, < client, question >, < client, answer >; then, the extracting unit recombines the binary knowledge structure into a ternary knowledge structure, wherein the ternary knowledge structure comprises < client, question, answer >, for example, a question and an answer corresponding to each other in the same session for the same client are taken as a ternary knowledge structure; then, the extracting unit clusters each element in the ternary knowledge structure to obtain the knowledge, wherein the structure of the knowledge includes < customer attribute, question, answer >, for example, the extracting unit abstracts the customer attributes of the customer, such as industry, region, scale, etc., clusters each element in the ternary knowledge structure based on the customer attributes to obtain a corresponding clustering result, and if three elements in the clustering result, namely, the customer, the question, and the answer, appear in the same session, the customer attribute, the question, and the corresponding answer corresponding to the customer are taken as one knowledge, thereby obtaining the knowledge with a knowledge structure of < customer attribute, question, answer >.
It will be understood by those skilled in the art that the foregoing manner of extracting knowledge is by way of example only and that other existing or future manners of extracting knowledge, such as may be applicable to the present invention, are intended to be included within the scope of the present invention and are hereby incorporated by reference. It will also be appreciated by those skilled in the art that the above-described customer attributes are exemplary only, and that other existing or future customer attributes, which may be suitable for use with the present invention, are also included within the scope of the present invention and are hereby incorporated by reference.
The updating means 205 creates or updates a corresponding knowledge base based on the knowledge.
Specifically, the updating device 205 stores the knowledge extracted by the knowledge extraction device 204 into the corresponding knowledge base, so as to implement the establishment or updating of the knowledge base, wherein the knowledge is stored in the knowledge base by the structure < customer attribute, question, answer >. The knowledge base may be a general database, preferably a lucene text index base.
Here, the knowledge base may be located in the recommendation apparatus 1, or may be located in a third-party device connected to the recommendation apparatus 1 through a network, and the updating apparatus 205 interacts with the knowledge base through the network, and when new knowledge is extracted, the new knowledge is stored in the knowledge base through the network.
The recommending device 1 extracts knowledge from the historical conversation records between each user and the client thereof, and stores the knowledge in a mode of < client attribute, question and answer >, so that when a user and the client make a call subsequently, the recommending answer is matched from the attribute information of the client to the user for answering, on one hand, the management of the knowledge is realized, on the other hand, the knowledge is recommended in real time according to the context, the system efficiency is improved, and the use experience of the user is improved.
Preferably (see fig. 1), the recommending device 1 further comprises a device for determining the priority of the recommended answer according to the occurrence frequency of the recommended answer and the corresponding current question in the knowledge base, hereinafter referred to as priority device (not shown); wherein the providing device 102 provides the recommended answer to the current user according to the priority.
Specifically, the recommended answer may have a certain priority, for example, each question or answer may appear more than once in the knowledge base, the priority device counts the number of times that the recommended answer and the current question corresponding to the recommended answer appear in the knowledge base according to the recommended answer and the current question corresponding to the recommended answer obtained by matching by the matching device 101, and determines the priority of the recommended answer, for example, the more the number of occurrences, the higher the priority is, the less the number of occurrences, the lower the priority is, here, the priority device may determine the priority according to the number of occurrences of the recommended answer or the number of occurrences of the current question, or may determine the priority by comprehensively considering both of them; subsequently, the providing device 102 provides the recommended answer to the current user according to the determined priority, for example, provides the recommended answer to the current user in a ranking of high to low priority.
Preferably, the knowledge base stores the number of times of occurrence of the question or answer in the entire knowledge base at the same time, for example, the knowledge is stored in the knowledge base in the structure of < customer attribute, question, number of times of occurrence of the question, answer, number of occurrence of answer >.
Here, according to the number of occurrences of the recommended answer and the current question corresponding to the recommended answer in the knowledge base, the recommending apparatus 1 may determine the priority of the recommended answer, and provide the recommended answer to the current user according to the priority, thereby further improving the user experience of the user.
More preferably (see fig. 1), the recommending device further comprises a device for adjusting the priority of the recommended answer according to the actual answer of the current user to the current question in the current session, hereinafter referred to as adjusting device (not shown).
Specifically, after providing the recommended answer obtained by matching by the matching device 101 to the user, the providing device 102 may select to answer the current question of the current client according to or with reference to the recommended answer, or may select not to answer according to the recommended answer, so that the adjusting device may obtain the actual answer of the current user to the current question in the current session, for example, the adjusting device obtains the actual answer of the current user to the current question in the current session in real time through interaction with a voice call device used by the current user; then, the adjusting device adjusts the priority of the recommended answer according to the actual answer, for example, the adjusting device compares the obtained actual answer with the recommended answer, if the obtained actual answer is consistent with the recommended answer, the current user is indicated to adopt the recommended answer, the adjusting device may increase the priority of the recommended answer, if the obtained actual answer is inconsistent with the recommended answer, the current user is indicated not to adopt the recommended answer, and the adjusting device may decrease the priority of the recommended answer.
Preferably, the adjusting means may adjust the priority of the recommended answer by counting the usage rate of the recommended answer by a large number of users, for example, a usage rate threshold value may be set, and when the usage rate of the recommended answer is lower than the usage rate threshold value, the priority of the recommended answer is lowered. Here, the usage threshold is, for example, preset by the system, and may be adjusted by the user according to the actual situation.
Here, the actual answer or the recommended answer may be in a text form or a speech form, and if the actual answer and the recommended answer are both in a speech form, the adjusting device may convert the actual answer and the recommended answer into text information by means of speech recognition or the like, and then compare the text information.
Here, "the actual answer" is identical to "the recommended answer" and "the identical" does not mean that the actual answer and the recommended answer are identical to each other, but may be considered to be identical as long as the similarity reaches a predetermined threshold, for example, for the actual answer and the recommended answer that have been converted into text information, if the similarity of the two reaches 90%, the actual answer and the recommended answer may be considered to be identical to each other. Here, the predetermined threshold is, for example, preset by the system, and may be adjusted by the user according to the actual situation.
Here, the recommending device 1 may also adjust the priority of the recommended answer according to whether the actual answer of the current user is consistent with the recommended answer, so that the actual feedback of the user is effectively utilized, and the use experience of the user is further improved.
Preferably (see fig. 1), the recommendation device 1 further comprises a device for modifying the attribute information of the current client in real time according to the question and answer of the current user and the current client in the current session, hereinafter referred to as modifying device (not shown).
Specifically, the current user and the current client can have several times of questions and answers in the current session, the current client can be a questioner or an answerer, in the question or answer of the current client, the attribute information of the current client, such as the industry, region, scale and the like of the current client, may be implied, the attribute information disclosed by these current clients in the current session may be in and out of the attribute information pre-stored in the system or known by the current user, the modifying means may thus obtain the question and answer of the current user and the current client in the current session, for example, and if the attribute information is different from the attribute information of the current client pre-stored in the system, the modifying device modifies the attribute information of the current client in real time.
Based on the modified attribute information, the matching device 101 may obtain a recommended answer corresponding to the current question in the current session for the current client matching again, or the matching device 101 may match the recommended answer corresponding to the modified attribute information when matching the next round of questions for the current client; or, if the current client has ended the current session, the modified attribute information may be used to match the corresponding recommended answer when the current client is in the next session.
Here, the recommendation device 1 modifies the attribute information of the current client in real time according to the question and answer of the client, so that the subsequent matching operation is more accurate, and the use experience of the user is further improved.
Preferably, the recommendation apparatus 1 further comprises means for acquiring the current session between the current user and the current client as a history session record for performing corresponding knowledge extraction, hereinafter referred to as second acquisition means (not shown).
Specifically, for the current session between the current user and the current client, it may also be used as a history session record for knowledge extraction, and the second obtaining device obtains the current session in real time, for example, through interaction with a voice communication device, for example, the second obtaining device obtains a current session each time the current user or the current client says one sentence, or, after the current user and the current client complete all questions and answers in the current session, the second obtaining device obtains all voices in the current session and uses the voices as a history session record, so as to perform corresponding knowledge extraction, for example, the current user and the current client perform knowledge extraction.
FIG. 3 illustrates a flow diagram of a method for session-based recommendation in accordance with another aspect of the subject innovation.
In step S301, the recommendation apparatus 1 matches and obtains a recommended answer corresponding to a current question in a current session from the knowledge base according to attribute information of a current client based on the current session between a current user and the current client, where a structure of knowledge in the knowledge base includes < client attribute, question, answer >.
Specifically, various types of knowledge are stored in the knowledge base, for example, the knowledge is stored in the knowledge base in a structure of < client attribute, question, answer >, and the knowledge is extracted from a large number of users according to the experience of the conversation with the clients thereof. For example, when a customer with a certain attribute asks a certain question, what the given answer is, the other users can refer to how to answer the customer's question conveniently by counting, sorting and storing the questions or answers in the knowledge base. The knowledge base can store various kinds of knowledge in a voice form or a character form, and can be located in the recommendation device 1 or in a third-party device connected with the recommendation device 1 through a network.
When a current user performs a current session with a current client thereof, for example, the current user is performing a voice call, in step S301, the recommending apparatus 1 matches and obtains a recommended answer corresponding to a current question in the current session from the knowledge base according to the attribute information of the current client, for example, in step S301, the recommending apparatus 1 first matches and obtains a client attribute corresponding to the attribute information and all knowledge corresponding to the client attribute in the knowledge base according to the attribute information of the current client, such as the attribute information of the industry, the region, the scale, and the like of the current client, and then matches and obtains a recommended answer corresponding to the question in the knowledge according to a current question asked by the current client in the current session.
Here, the attribute information of the current client is, for example, known to the current user, for example, known before the current user performs the current session, or may be attribute information analyzed or adjusted in real time by the recommendation apparatus 1.
Here, in step S301, the recommendation device 1 may also convert the current question in the current session into text information through voice recognition, and perform matching search in the knowledge base according to the text information.
It should be understood by those skilled in the art that the above-mentioned method for matching recommended answers is only an example, and other methods for matching recommended answers that are currently available or that may be later developed, such as being applicable to the present invention, are also included in the scope of the present invention and are also included herein by reference.
In step S302, the recommendation device 1 provides the recommendation answer to the current user.
Specifically, for the recommended answer obtained by matching with the recommending device 1 in step S301, in step S302, the recommending device 1 provides the recommended answer to the current user for the current user to select, for example, through a presentation manner of voice, text, or other convention. For example, when the current user uses a wearable device to perform a current session with the current client, in step S302, the recommending apparatus 1 interacts with the wearable device, and provides the recommended answer matched in step S301 to the current user via the wearable device. Preferably, for the current question in the current session, if the recommending apparatus 1 matches a plurality of recommended answers in step S301, in step S302, the recommending apparatus 1 may provide the plurality of recommended answers to the user randomly or in a certain order.
It should be understood by those skilled in the art that the above-mentioned manner for providing recommended answers is only an example, and other manners for providing recommended answers that are currently available or that may come out later, such as may be applicable to the present invention, are also included in the scope of the present invention and are herein incorporated by reference.
The recommendation device 1 matches and obtains a recommendation answer corresponding to a current question in a current session from a knowledge base according to attribute information of a current client based on the current session between the current user and the current client, and provides the recommendation answer to the current user, wherein the structure of knowledge in the knowledge base comprises < client attribute, question and answer >, so that the current knowledge is effectively utilized, the current user can conveniently answer the inquiry of the current client, and the use experience of the user is improved.
FIG. 4 shows a flow diagram of a method for session-based recommendation, according to another embodiment of the present invention.
The method further comprises step S403: acquiring historical session records between each user and the client of the user; step S404: extracting knowledge from the historical conversation records to obtain corresponding knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers; and step S405: and establishing or updating the knowledge base according to the knowledge. Steps S401 and S402 are the same as or substantially the same as the corresponding steps shown in fig. 3, and thus are not described herein again and are included herein by way of reference.
In step S403, the recommendation apparatus 1 acquires a history of session between each user and its client.
Specifically, when the user communicates with the client, the content of the communication may be recorded, for example, when the user makes a call with the client, the recommendation device 1 records the content of the voice call by recording through the voice call device used by the user, and in step S403, periodically or in real time, for example, by interacting with each voice call device, acquires a plurality of historical conversation records between each user and the client.
It should be understood by those skilled in the art that the above-mentioned manner of obtaining historical session records is only an example, and other existing or future manners of obtaining historical session records, such as those that may be applicable to the present invention, are also included within the scope of the present invention and are incorporated herein by reference.
In step S404, the recommending apparatus 1 extracts knowledge from the historical conversation record to obtain corresponding knowledge, where the structure of the knowledge includes < customer attribute, question, answer >.
Specifically, for the historical conversation record obtained in step S403, in step S404, the recommendation device 1 performs knowledge extraction on the historical conversation record, for example, first converts the historical conversation record in the voice form into a historical conversation record in the text form, and then extracts knowledge from the historical conversation record in the text form, where the structure of the knowledge includes < customer attribute, question, answer >, and for example, the question or answer corresponding to the customer is classified according to the customer attribute corresponding to different customers, and then the question and answer belonging to one conversation are taken as one knowledge, so as to obtain the knowledge with the structure of < customer attribute, question, answer >.
Preferably, the step S404 includes a sub-step S404a (not shown) and a sub-step S404b (not shown). In sub-step S404a, the recommendation device 1 converts the historical conversation record into corresponding conversation text information; in sub-step S404b, the recommending apparatus 1 performs knowledge extraction on the session text information to obtain the knowledge, wherein the structure of the knowledge includes < customer attribute, question, answer >.
Specifically, for the history session record in the voice form acquired in step S403, in sub-step S404a, the recommendation device 1 converts the history session record in the voice form into corresponding session text information, for example, by using a voice recognition technology, a character conversion technology, or the like; further, in sub-step S404a, the recommendation device 1 may first perform pre-processing on the historical conversation record acquired in step S403, such as determining the language family of the historical conversation record, determining the front and back nasal sounds of the historical conversation record, and so on, to obtain pre-processed information; and determining corresponding session text information according to the historical session record or further combining the user related information and the preprocessing information.
Subsequently, in sub-step S404b, the recommending apparatus 1 extracts knowledge of the conversation text information converted in sub-step S404a to obtain the knowledge, for example, classifies questions or answers corresponding to different customers according to their corresponding customer attributes, and uses the questions and answers belonging to the same conversation as one knowledge, thereby obtaining knowledge having the structure of < customer attributes, questions, answers >.
More preferably, in sub-step S404b, the recommendation device 1 abstracts the session text information into a binary knowledge structure, wherein the binary knowledge structure comprises < user or client, say >; recombining the binary knowledge structure into a ternary knowledge structure, wherein the ternary knowledge structure comprises < client, question, answer >; and clustering each element in the ternary knowledge structure to obtain the knowledge, wherein the structure of the knowledge comprises client attributes, questions and answers.
Specifically, after the recommender 1 converts the historical conversation record into the conversation text information in the sub-step S404a, in the sub-step S404b, the recommender 1 abstracts the conversation text information into a binary knowledge structure, for example, by means of classification, wherein the binary knowledge structure includes < user or client, say >, where the say may be a question or an answer, that is, in the sub-step S404b, the recommender 1 classifies the conversation text information into < user, question >, < user, answer >, < client, question >, < client, answer >; subsequently, the recommending apparatus 1 regroups the binary knowledge structure into a ternary knowledge structure, wherein the ternary knowledge structure comprises < client, question, answer >, for example, a question and an answer corresponding to each other in the same session for the same client are taken as a ternary knowledge structure; then, the recommending apparatus 1 clusters each element in the ternary knowledge structure to obtain the knowledge, wherein the structure of the knowledge includes < customer attribute, question, answer >, for example, in sub-step S404b, the recommending apparatus 1 abstracts the customer attributes such as industry, region, scale, etc. of the customer, clusters each element in the ternary knowledge structure based on the customer attributes to obtain a corresponding clustering result, and if three elements, i.e., customer, question, and answer, in the clustering result appear in the same session, takes the customer attribute, question, and corresponding answer corresponding to the customer as one knowledge, thereby obtaining the knowledge whose knowledge structure is < customer attribute, question, answer >.
It will be understood by those skilled in the art that the foregoing manner of extracting knowledge is by way of example only and that other existing or future manners of extracting knowledge, such as may be applicable to the present invention, are intended to be included within the scope of the present invention and are hereby incorporated by reference. It will also be appreciated by those skilled in the art that the above-described customer attributes are exemplary only, and that other existing or future customer attributes, which may be suitable for use with the present invention, are also included within the scope of the present invention and are hereby incorporated by reference.
In step S405, the recommendation device 1 creates or updates a corresponding knowledge base according to the knowledge.
Specifically, in step S405, the recommendation device 1 stores the knowledge extracted in step S404 into the corresponding knowledge base, so as to establish or update the knowledge base, where the knowledge is stored in the structure < customer attribute, question, answer >. The knowledge base may be a general database, preferably a lucene text index base.
Here, the knowledge base may be located in the recommendation apparatus 1 or in a third-party device connected to the recommendation apparatus 1 through a network, and in step S405, the recommendation apparatus 1 interacts with the knowledge base through the network, and when new knowledge is extracted, the new knowledge is stored in the knowledge base through the network.
The recommending device 1 extracts knowledge from the historical conversation records between each user and the client thereof, and stores the knowledge in a mode of < client attribute, question and answer >, so that when a user and the client make a call subsequently, the recommending answer is matched from the attribute information of the client to the user for answering, on one hand, the management of the knowledge is realized, on the other hand, the knowledge is recommended in real time according to the context, the system efficiency is improved, and the use experience of the user is improved.
Preferably (see fig. 3), the method further comprises step S306 (not shown): determining the priority of the recommended answer according to the occurrence times of the recommended answer and the current question corresponding to the recommended answer in the knowledge base; in step S302, the recommending apparatus 1 provides the recommended answer to the current user according to the priority.
Specifically, the recommended answer may have a certain priority, for example, each question or answer may appear more than once in the knowledge base, in step S306, the recommending apparatus 1 counts the number of times that the recommended answer and the current question corresponding to the recommended answer appear in the knowledge base according to the recommended answer and the current question corresponding to the recommended answer obtained by matching in step S301, and determines the priority of the recommended answer, for example, the more the number of occurrences, the higher the priority is, the less the number of occurrences, the lower the priority is, in step S306, the recommending apparatus 1 may determine the priority according to the number of times that the recommended answer appears or the number of times that the current question appears, or may determine the priority by comprehensively considering both of the numbers; subsequently, in step S302, the recommending apparatus 1 provides the recommended answer to the current user according to the determined priority, for example, the recommended answer is provided to the current user in a ranking of high to low priority.
Preferably, the knowledge base stores the number of times of occurrence of the question or answer in the entire knowledge base at the same time, for example, the knowledge is stored in the knowledge base in the structure of < customer attribute, question, number of times of occurrence of the question, answer, number of occurrence of answer >.
Here, according to the number of occurrences of the recommended answer and the current question corresponding to the recommended answer in the knowledge base, the recommending apparatus 1 may determine the priority of the recommended answer, and provide the recommended answer to the current user according to the priority, thereby further improving the user experience of the user.
More preferably (see fig. 3), the method further comprises step S307 (not shown): and adjusting the priority of the recommended answer according to the actual answer of the current user to the current question in the current session.
Specifically, in step S302, after providing the recommended answer matched in step S301 to the user, the recommending apparatus 1 may select to answer the current question of the current client according to or with reference to the recommended answer, or may select not to answer according to the recommended answer, so in step S307, the recommending apparatus 1 may obtain the actual answer of the current user to the current question in the current session, for example, in step S307, the recommending apparatus 1 obtains the actual answer of the current user to the current question in the current session in real time through interaction with a voice call apparatus used by the current user; subsequently, in step S307, the recommending apparatus 1 adjusts the priority of the recommended answer according to the actual answer, for example, in step S307, the recommending apparatus 1 compares the obtained actual answer with the recommended answer, if the obtained actual answer is consistent with the recommended answer, the recommending apparatus 1 may increase the priority of the recommended answer, and if the obtained actual answer is inconsistent with the recommended answer, the recommending apparatus 1 may not adopt the recommended answer, and the recommending apparatus 1 may decrease the priority of the recommended answer.
Preferably, in step S307, the recommending apparatus 1 may further adjust the priority of the recommended answer by counting the usage rate of the recommended answer by a large number of users, for example, a usage rate threshold may be set, and when the usage rate of the recommended answer is lower than the usage rate threshold, the priority of the recommended answer is lowered. Here, the usage threshold is, for example, preset by the system, and may be adjusted by the user according to the actual situation.
Here, the actual answer or the recommended answer may be in a text form or a voice form, and if the actual answer and the recommended answer are both in a voice form, in step S307, the recommending apparatus 1 may convert the actual answer and the recommended answer into text information by means of voice recognition or the like, and then compare the text information.
Here, "the actual answer" is identical to "the recommended answer" and "the identical" does not mean that the actual answer and the recommended answer are identical to each other, but may be considered to be identical as long as the similarity reaches a predetermined threshold, for example, for the actual answer and the recommended answer that have been converted into text information, if the similarity of the two reaches 90%, the actual answer and the recommended answer may be considered to be identical to each other. Here, the predetermined threshold is, for example, preset by the system, and may be adjusted by the user according to the actual situation.
Here, the recommending device 1 may also adjust the priority of the recommended answer according to whether the actual answer of the current user is consistent with the recommended answer, so that the actual feedback of the user is effectively utilized, and the use experience of the user is further improved.
Preferably (see fig. 3), the method further comprises step S308 (not shown): and modifying the attribute information of the current client in real time according to the question and answer of the current user and the current client in the current session.
Specifically, the current user and the current client may have several questions and answers in the current session, the current client may be a questioner or an answerer, in the questions or answers of the current client, the attribute information of the current client may be implied, such as the attribute information of the industry, the region, the scale, and the like of the current client, and the attribute information disclosed by the current client in the current session may come in or go out with the attribute information pre-stored in the system or known by the current client, so in step S308, the recommending apparatus 1 may obtain the questions and answers of the current user and the current client in the current session, for example, by interacting with the voice call apparatus, obtain the questions and answers, and extract the attribute information of the current client from the questions and answers, if the attribute information is different from the attribute information of the current client pre-stored in the system, the recommendation apparatus 1 modifies the attribute information of the current client in real time in step S308.
Based on the modified attribute information, in step S301, the recommending apparatus 1 may obtain a recommended answer corresponding to the current question in the current session for the current customer matching again, or, in step S301, the recommending apparatus 1 may match the recommended answer corresponding to the modified attribute information when matching the next round of questions for the current customer; or, if the current client has ended the current session, the modified attribute information may be used to match the corresponding recommended answer when the current client is in the next session.
Here, the recommendation device 1 modifies the attribute information of the current client in real time according to the question and answer of the client, so that the subsequent matching operation is more accurate, and the use experience of the user is further improved.
Preferably (see fig. 3), the method further comprises step S309 (not shown): and acquiring the current session between the current user and the current client as a historical session record so as to extract corresponding knowledge.
Specifically, for the current session between the current user and the current client, which may also be used as a history session record for knowledge extraction, in step S309, the recommendation apparatus 1 acquires the current session in real time, for example, through interaction with a voice communicator, for example, whenever the current user or the current client finishes a sentence, in step S309, the recommendation apparatus 1 acquires the sentence, or, after the current user and the current client complete all questions and answers in the current session, in step S309, the recommendation apparatus 1 acquires all voices in the current session and records the voices as history sessions, so as to perform corresponding knowledge extraction, for example, to submit to the recommendation apparatus 1 for knowledge extraction.
It is noted that the present invention may be implemented in software and/or in a combination of software and hardware, for example, the various means of the invention may be implemented using Application Specific Integrated Circuits (ASICs) or any other similar hardware devices. In one embodiment, the software program of the present invention may be executed by a processor to implement the steps or functions described above. Also, the software programs (including associated data structures) of the present invention can be stored in a computer readable recording medium, such as RAM memory, magnetic or optical drive or diskette and the like. Further, some of the steps or functions of the present invention may be implemented in hardware, for example, as circuitry that cooperates with the processor to perform various steps or functions.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.
Claims (15)
1. A method for making recommendations based on a session, wherein the method comprises the steps of:
based on a current session between a current user and a current client thereof, matching and obtaining a recommended answer corresponding to a current question in the current session from a knowledge base according to the attribute information of the current client;
providing the recommended answer to the current user;
the structure of knowledge in the knowledge base comprises client attributes, questions and answers, wherein the client corresponding to the client attributes, the questions and the answers are three elements appearing in the same session in a clustering result, the clustering result is obtained after all elements in a ternary knowledge structure obtained based on historical session records are clustered, and the ternary knowledge structure comprises the client, the questions and the answers.
2. The method of claim 1, wherein the knowledge base is established or updated by:
acquiring the historical session records between each user and the client thereof;
converting the historical conversation record into corresponding conversation text information;
extracting knowledge from the session text information to obtain the ternary knowledge structure;
and clustering each element in the ternary knowledge structure respectively to obtain the knowledge.
3. The method of claim 2, wherein the step of extracting knowledge from the session text information to obtain the ternary knowledge structure comprises:
abstracting the session text information into a binary knowledge structure, wherein the binary knowledge structure comprises < user or client, say >;
and recombining the binary knowledge structure into the ternary knowledge structure.
4. The method of any of claims 1 to 3, wherein the method further comprises:
determining the priority of the recommended answer according to the occurrence times of the recommended answer and the current question corresponding to the recommended answer in the knowledge base;
wherein the step of providing the recommended answer to the current user comprises:
and providing the recommended answer to the current user according to the priority.
5. The method of claim 4, wherein the method further comprises:
and adjusting the priority of the recommended answer according to the actual answer of the current user to the current question in the current session.
6. The method of any of claims 1-3, 5, wherein the method further comprises:
and modifying the attribute information of the current client in real time according to the question and answer of the current user and the current client in the current session.
7. The method of any of claims 1-3, 5, wherein the method further comprises:
and acquiring the current session between the current user and the current client as a historical session record so as to extract corresponding knowledge.
8. A recommendation apparatus that makes a recommendation based on a session, wherein the recommendation apparatus comprises:
the device is used for matching and obtaining a recommended answer corresponding to a current question in the current conversation from a knowledge base according to the attribute information of the current client based on the current conversation between the current user and the current client;
means for providing the recommended answer to the current user;
the structure of knowledge in the knowledge base comprises client attributes, questions and answers, wherein the client corresponding to the client attributes, the questions and the answers are three elements appearing in the same session in a clustering result, the clustering result is obtained after all elements in a ternary knowledge structure obtained based on historical session records are clustered, and the ternary knowledge structure comprises the client, the questions and the answers.
9. The recommendation device of claim 8, further comprising:
means for obtaining said historical session records between respective users and their clients;
means for converting the historical conversation record into corresponding conversation text information;
a unit for extracting knowledge from the session text information to obtain the ternary knowledge structure;
and clustering each element in the ternary knowledge structure respectively to obtain the knowledge unit.
10. The recommendation apparatus of claim 9, wherein the means for knowledge extraction of the session text information to obtain the ternary knowledge structure is configured to:
abstracting the session text information into a binary knowledge structure, wherein the binary knowledge structure comprises < user or client, say >;
and recombining the binary knowledge structure into the ternary knowledge structure.
11. The recommendation device according to any one of claims 8 to 10, further comprising:
means for determining a priority of the recommended answer based on the number of occurrences of the recommended answer and its corresponding current question in the knowledge base;
wherein the means for providing the recommended answer to the current user is to:
and providing the recommended answer to the current user according to the priority.
12. The recommendation device of claim 11, further comprising:
means for adjusting a priority of the recommended answer according to an actual answer of the current user to the current question in the current session.
13. The recommendation device according to any one of claims 8 to 10, 12, further comprising:
and modifying the attribute information of the current client in real time according to the question and answer of the current user and the current client in the current session.
14. The recommendation device according to any one of claims 8 to 10, 12, further comprising:
and the device is used for acquiring the current session between the current user and the current client as a historical session record so as to extract corresponding knowledge.
15. A computer device, the computer device comprising:
one or more processors;
a memory for storing one or more computer programs;
the one or more computer programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610798166.1A CN107784033B (en) | 2016-08-31 | 2016-08-31 | Method and device for recommending based on session |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610798166.1A CN107784033B (en) | 2016-08-31 | 2016-08-31 | Method and device for recommending based on session |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107784033A CN107784033A (en) | 2018-03-09 |
| CN107784033B true CN107784033B (en) | 2021-10-22 |
Family
ID=61451782
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610798166.1A Active CN107784033B (en) | 2016-08-31 | 2016-08-31 | Method and device for recommending based on session |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107784033B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110874781A (en) * | 2018-09-03 | 2020-03-10 | 上海汽车集团股份有限公司 | Automobile matching method and system |
| CN109460504A (en) * | 2018-09-21 | 2019-03-12 | 广州神马移动信息科技有限公司 | The answer main body recommended method and its device, electronic equipment, computer-readable medium of answer are reserved in Knowledge Community |
| CN109408678A (en) * | 2018-09-21 | 2019-03-01 | 广州神马移动信息科技有限公司 | Question and answer exchange method and its device |
| CN109446411A (en) * | 2018-09-21 | 2019-03-08 | 广州神马移动信息科技有限公司 | Answer method and device thereof, electronic equipment, computer-readable medium are reserved in Knowledge Community |
| CN109766425A (en) * | 2018-12-29 | 2019-05-17 | 重庆木舌科技有限公司 | The online system of campus legal advice in real time |
| CN110203154B (en) * | 2019-05-29 | 2022-09-30 | 阿波罗智联(北京)科技有限公司 | Recommendation method and device for vehicle functions, electronic equipment and computer storage medium |
| CN111581363B (en) * | 2020-04-30 | 2023-08-29 | 北京百度网讯科技有限公司 | Knowledge extraction method, device, equipment and storage medium |
| CN112015789A (en) * | 2020-09-01 | 2020-12-01 | 中国银行股份有限公司 | Knowledge base retrieval method and device, electronic equipment and computer storage medium |
| CN112818225A (en) * | 2021-01-27 | 2021-05-18 | 上海明略人工智能(集团)有限公司 | Display method and device of pushed data |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103081491A (en) * | 2010-08-27 | 2013-05-01 | 索尼公司 | Transmitting device and method, receiving device and method, and transmitting-receiving system |
| CN103455535A (en) * | 2013-05-08 | 2013-12-18 | 深圳市明唐通信有限公司 | Method for establishing knowledge base based on historical consultation data |
| CN105335400A (en) * | 2014-07-22 | 2016-02-17 | 阿里巴巴集团控股有限公司 | Method and apparatus for obtaining answer information for questioning intention of user |
| CN105591882A (en) * | 2015-12-10 | 2016-05-18 | 北京中科汇联科技股份有限公司 | Method and system for mixed customer services of intelligent robots and human beings |
| CN105824970A (en) * | 2016-04-12 | 2016-08-03 | 华南师范大学 | Robot interaction method and system based on big data knowledge base and user feedback |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150309988A1 (en) * | 2014-04-29 | 2015-10-29 | International Business Machines Corporation | Evaluating Crowd Sourced Information Using Crowd Sourced Metadata |
-
2016
- 2016-08-31 CN CN201610798166.1A patent/CN107784033B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103081491A (en) * | 2010-08-27 | 2013-05-01 | 索尼公司 | Transmitting device and method, receiving device and method, and transmitting-receiving system |
| CN103455535A (en) * | 2013-05-08 | 2013-12-18 | 深圳市明唐通信有限公司 | Method for establishing knowledge base based on historical consultation data |
| CN105335400A (en) * | 2014-07-22 | 2016-02-17 | 阿里巴巴集团控股有限公司 | Method and apparatus for obtaining answer information for questioning intention of user |
| CN105591882A (en) * | 2015-12-10 | 2016-05-18 | 北京中科汇联科技股份有限公司 | Method and system for mixed customer services of intelligent robots and human beings |
| CN105824970A (en) * | 2016-04-12 | 2016-08-03 | 华南师范大学 | Robot interaction method and system based on big data knowledge base and user feedback |
Non-Patent Citations (1)
| Title |
|---|
| 基于人工智能的湖南移动在线客服研究;令妍;《万方学位论文电子库》;20150817;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107784033A (en) | 2018-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107784033B (en) | Method and device for recommending based on session | |
| JP7464098B2 (en) | Electronic conference system | |
| US11004013B2 (en) | Training of chatbots from corpus of human-to-human chats | |
| CN107680019B (en) | Examination scheme implementation method, device, equipment and storage medium | |
| CN106685916B (en) | Intelligent device and method for electronic conference | |
| JP6718828B2 (en) | Information input method and device | |
| US11003863B2 (en) | Interactive dialog training and communication system using artificial intelligence | |
| CN107153965A (en) | A kind of intelligent customer service solution of multiple terminals | |
| CN107451110A (en) | A kind of method, apparatus and server for generating meeting summary | |
| CN107436916B (en) | Intelligent answer prompting method and device | |
| EP4297030A2 (en) | Polling questions for a conference call discussion | |
| US20180075354A1 (en) | Method and system for using existing models in connection with new model development | |
| CN116595150A (en) | Dialogue recommendation method, device, equipment and storage medium | |
| CN116501960B (en) | Content retrieval method, device, equipment and medium | |
| CN112418779A (en) | Online self-service interviewing method based on natural language understanding | |
| CN114254158A (en) | Video generation method and device, and neural network training method and device | |
| CN111723180A (en) | Interviewing method and device | |
| CN111858560B (en) | Financial data automatic testing and monitoring system based on data warehouse | |
| CN112199474A (en) | Voice customer service method and system | |
| CN110311943A (en) | The inquiry of data and methods of exhibiting in a kind of electric power enterprise big data platform | |
| CN113505606B (en) | Training information acquisition method and device, electronic equipment and storage medium | |
| CN115831125A (en) | Speech recognition method, device, equipment, storage medium and product | |
| CN114240250A (en) | Intelligent management method and system for vocational evaluation | |
| CN111695360A (en) | Semantic analysis method and device, electronic equipment and storage medium | |
| US20240112197A1 (en) | Artificial Intelligence Based Topic Detection |
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 |