CN112330387B - Virtual broker applied to house watching software - Google Patents
Virtual broker applied to house watching software Download PDFInfo
- Publication number
- CN112330387B CN112330387B CN202011050180.6A CN202011050180A CN112330387B CN 112330387 B CN112330387 B CN 112330387B CN 202011050180 A CN202011050180 A CN 202011050180A CN 112330387 B CN112330387 B CN 112330387B
- Authority
- CN
- China
- Prior art keywords
- house
- information
- module
- unit
- client
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0623—Item investigation
- G06Q30/0625—Directed, with specific intent or strategy
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0641—Shopping interfaces
- G06Q30/0643—Graphical representation of items or shoppers
-
- Y—GENERAL 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention provides a virtual broker applied to house watching software, comprising: the system comprises an information receiving module, an information processing module, a storage module, a communication module, a house recommending module, a client recommending module and an intelligent answering module; the system comprises a storage module, a storage module and a query module, wherein the storage module is used for storing a resource library and house introduction information, the resource library is used for storing house source information, house introduction information and customer information, the knowledge library is used for storing a map and a problem set, the map is formed by integrating information in the resource library, and the problem set is a historical question-answer set of a customer. The method and the system realize house source recommendation for the clients through the information receiving module and the information processing module and the house recommendation module to recommend high-quality clients; and based on the map and the intelligent answer module, the questions of the clients are answered, so that the clients can more intuitively know the basic conditions and the peripheral information of the house, the house can be deeply understood without on-site house watching, and the time cost is saved.
Description
Technical Field
The invention relates to the technical field of computer communication, in particular to a virtual broker applied to house watching software.
Background
In a property marketing campaign, a traditional house-watching app can only see basic information and house interior profiles, and many other information needs to be communicated through a broker, and finally, the house-watching app can watch a house on site with a large probability. From the perspective of customers, the house can directly communicate with the house in real time without communicating with a broker, so that the house itself can tell the customers about own basic conditions and site information, the house can be seen without reservation, a lot of time can be saved, the cost can be saved, and a lot of time conflict problems can be further solved, so that the house is very suitable for the current fast-paced life.
Disclosure of Invention
Based on this, it is necessary to provide a virtual broker applied to house-watching software in view of the above technical problems.
A virtual broker for use in house-seeing software, comprising: the system comprises an information receiving module, an information processing module, a storage module, a communication module, a house recommending module, a client recommending module and an intelligent answering module; the system comprises a storage module, a storage module and a query module, wherein the storage module is used for storing a resource library and a knowledge library, the resource library is used for storing house source information, house introduction information and customer information, the knowledge library is used for storing a map and a problem set, the map is formed by integrating information in the resource library, and the problem set is a historical question-answer set of a customer.
In one embodiment, the house source information at least comprises house source house type, house address and peripheral information; the house introduction information at least comprises house official explanation information and house official evaluation information; the client information at least comprises client question and answer information, client browsing record information and client house evaluation information.
In one embodiment, the information receiving module is configured to receive various information including house source information, house introduction information, and customer information.
In one embodiment, the information processing module is configured to integrate information in the repository into a map and a problem set in the knowledge base.
In one embodiment, the house recommendation module includes: the system comprises a calculation and analysis submodule and a house recommendation submodule, wherein the calculation and analysis submodule comprises a model building unit, a similarity calculation unit, a neighbor iteration unit and a scoring prediction unit, wherein: the model building unit is used for building a preference matrix model of a user about a house source by utilizing the information in the resource library; the similarity calculation unit is used for calculating the similarity among the users by utilizing cosine similarity based on the preference matrix model to obtain a preference similarity matrix; the neighbor iteration unit is used for obtaining new similarity among users according to the attenuation ratio and the attenuation weight based on the similarity of each user, and iterating the neighbor set according to a neighbor set formed by the new similarity to obtain a target neighbor set; the scoring prediction unit is used for predicting the scoring of the user about the house source according to the client information based on the target neighbor set; and the house recommending submodule is used for sorting the scores according to the size and selecting the top three house sources as target house sources to recommend to the customer.
In one embodiment, the client recommendation module includes: a weight setting unit, a weight updating unit, a quality calculating unit and a client recommending unit, wherein: the weight setting unit is used for setting weights W1 and W2 for browsing duration T and problem consultation of clients respectively, sequencing questions and answers in the problem set according to the attention degree, and setting different weights W; the weight updating unit is used for updating the attention degree of the problem in the problem set according to the client information of the client and updating the weight w; the quality calculation unit is used for calculating the quality of the client according to the weights W1, W2 and W; the client recommending unit is used for taking the client with the quality larger than the threshold value as a target client according to the preset threshold value and pushing the client to the cloud through the communication module.
In one embodiment, the client recommending unit is further configured to push the target client to the knowledge base, and update the map.
In one embodiment, the intelligent answer module comprises: the system comprises a question analysis sub-module, an information retrieval sub-module and an answer generation sub-module, wherein the question analysis sub-module comprises an entity obtaining unit and a purpose obtaining unit, and the question analysis sub-module comprises a target obtaining unit, wherein: the entity obtaining unit is used for carrying out word segmentation and entity identification processing on the questioning of the client to obtain an entity of the questioning; the objective obtaining unit is used for carrying out objective identification and problem classification processing on the client questions and obtaining the objectives of the questions; the information retrieval submodule is used for inquiring the knowledge base according to the entity and the purpose and obtaining an initial inquiry result; the answer generation submodule is used for comparing and sequencing the initial query results, selecting an optimal answer, generating a target answer in a natural sentence form from the optimal answer, and feeding the target answer back to the client.
In one embodiment, the intelligent answer module further comprises a natural language generation sub-module comprising a content determination unit, a structure determination unit, and a language generation unit, wherein: the content determining unit is used for taking the entity and the target answer as statement key information; the structure determining unit is used for determining a sentence structure based on the problem set and according to the purpose; the language generating unit is used for combining the statement key information according to the statement structure so as to obtain a target answer in a natural statement form.
The virtual broker applied to house watching software can follow up the situation of the client in real time through the information receiving module and the information processing module, and can respectively realize house source recommendation to the client, and recommend high-quality client to customer service through the house recommendation module, the client recommendation module and the intelligent answer module, and answer the client questions based on the atlas in the knowledge base, so that the client can answer questions and answers, the basic situation and the peripheral information of the house are more intuitively known, the house can be deeply known without site house watching, and the time cost is saved.
Drawings
FIG. 1 is an application scenario diagram of a virtual broker applied to house-looking software in one embodiment;
FIG. 2 is a block diagram of a virtual broker applied to house-looking software in one embodiment;
FIG. 3 is a block diagram of the architecture of a house recommendation module in one embodiment;
FIG. 4 is a block diagram of the architecture of a client recommendation module in one embodiment;
FIG. 5 is a block diagram of the architecture of the intelligent answer module in one embodiment;
FIG. 6 is a block diagram of the computing and analysis sub-module in one embodiment;
FIG. 7 is a block diagram of the problem analysis sub-module in one embodiment;
FIG. 8 is a block diagram of the natural language generation sub-module, in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail by the following detailed description with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The virtual broker applied to house watching software can be applied to an application environment shown in fig. 1. The house watching software is installed on the terminal 1, the virtual broker 11 is applied in the house watching software, and the clients 12 can conduct questions and answers with the virtual broker 11 on the house watching software, the communication module of the virtual broker 11 is communicated with the server 2, so that high-quality clients can be uploaded to the server 2 through the communication module, and the server 2 can send the clients to needed customer service or store the clients in the server 2. The terminal 1 may be, but not limited to, various personal computers, notebook computers, smartphones, tablet computers and portable wearable devices, and the server 2 may be implemented by a stand-alone server or a server cluster composed of a plurality of servers.
In one embodiment, as shown in fig. 2, there is provided a virtual broker 200 applied to house watching software, including an information receiving module 210, an information processing module 220, a storage module 230, a communication module 240, a house recommendation module 250, a client recommendation module 260, and an intelligent answer module 270, wherein a resource library and a knowledge library are stored in the storage module 230, the resource library is used for storing house source information, house introduction information, and client information, the knowledge library is used for storing a map and a question set, the map is formed by integrating information in the resource library, and the question set is a historical question-answer set of a client. The house source information at least comprises house source house types, house addresses and peripheral information; the house introduction information at least comprises house official explanation information and house official evaluation information; the client information includes at least client question-answer information, client browsing record information, and client house evaluation information. Wherein the information receiving module 210 is configured to receive various information, where the information includes house source information, house introduction information, and customer information; the information processing module 220 is configured to integrate information in the repository into a atlas and a problem set in the knowledge base.
Specifically, the main implementation principle of the virtual broker is: information retrieval is mainly divided into two modules: creating a knowledge base, and man-machine questions (retrieval). Creating a knowledge base: a knowledge map is established as a knowledge base of the virtual broker by using house source information, peripheral information of house source address positioning, broker explanation, old customer evaluation, browsing records, historical question-answer records and the like. Man-machine question answering: when a customer presents a question, the best answer is given through question analysis. APP scene: a floating button arranged on the top layer is arranged in the APP, and the button carries client information and the current position. After the client enters the APP, clicking the suspension button wakes up a virtual broker, can carry out voice dialogue like chatting with a property broker in reality, and after the client inquires about information about a house source, the broker can give corresponding answers and records the questions of the client. In addition, whether or not the client wakes up, the virtual broker records browsing information of the client and updates the information to the knowledge base.
Firstly, an information resource library is established, wherein the information of the resource library comprises three types, namely house source information, house introduction information and customer information. The house source information comprises all house source information under a developer and peripheral information of associated address positioning; house introduction information includes comments and evaluations of houses by house property brokers; the client information is the dialogue between the client and the virtual broker, and the browsing records of the client and comments on the house source.
The information processing module 220 is then used to integrate the information in the repository into a map and problem set in the knowledge base, initialize the knowledge base and generate a "map" when there is a new house entry, detect the surrounding information at regular time, and update the knowledge base if a change in information is detected ("corresponding data node in map"). The information resources that have been obtained are extracted into key information, and a "map" is built for each house source using the entities (house sources, addresses, customers, etc.), attributes (house types, perimeter, evaluation, customer quality, etc.). The knowledge graph is obtained by associating the maps of all the house sources by using the relation (associated clients, intention clients and the like), and the graph and the problem set jointly form a knowledge base. If a new market is opened around, a new node is generated and is associated with a corresponding house source; for example, the user browses or inquires the information related to the house source, and the association weight of the house source is updated according to the browsing time length of the user or the prediction intention of the consulting problem and the like, if a new visiting client is, a new connection is established to be associated with the house source.
In the above embodiment, a virtual broker applied to house watching software is provided, the situation of a customer is followed in real time through the information receiving module 210 and the information processing module 220, and house source recommendation is performed on the customer through the house recommending module 250, the customer recommending module 260 and the intelligent answering module 270, and high-quality customer is recommended to customer service, and the customer questioning can be answered based on the atlas in the knowledge base, so that the customer can know the basic situation and the peripheral information of the house more intuitively, and the house can be known deeply without site watching, thereby saving the time cost.
In one embodiment, as shown in fig. 3 and 6, the house recommendation module 250 includes: a calculation analysis sub-module 251 and a house recommendation sub-module 252, the calculation analysis sub-module 251 including a model building unit 251A, a similarity calculation unit 251B, a neighbor iteration unit 251C, and a score prediction unit 251D, wherein:
the model building unit 251A is configured to build a preference matrix model of the user about the house source by using information in the resource library;
the similarity calculation unit 251B is configured to calculate, based on the preference matrix model, similarity between each user by using cosine similarity, to obtain a preference similarity matrix;
the neighbor iteration unit 251C is configured to obtain new similarities between users according to the attenuation ratio and the attenuation weight based on the similarities of the users, and iterate the neighbor set according to a neighbor set formed by the new similarities to obtain a target neighbor set;
the scoring prediction unit 251D is configured to predict a score of the user about the house source according to the client information based on the target neighbor set;
the house recommendation sub-module 252 is configured to sort the scores by size and select the top three house sources as target house sources for recommendation to the customer.
Specifically, first, a customer preference analysis is performed, and the following sets are established according to information in a database:
house Source set I= { I 1 ，I 2 ,...,I n }
House Source tag set L= { L 1 ，l 2 ,...,l g }
User set u= { U 1 ，U 2 ,...,U m }
Analyzing the evaluation of the customer on the house source, converting the evaluation into a scoring matrix R, and calculating a preference vector Q= { Q of the customer according to the label of the house source browsed (implicit) by the customer and the house source evaluated (explicit) by the customer and the score l1 ，Q l2 ,...,Q lk }，Q lk The preference degree of a certain type of house source tag is calculated, and then preference vectors of m users can be calculated to obtain a preference matrix model.
Then, in the similarity calculation unit 251B, a similarity S (U) between users can be calculated using cosine similarity with respect to the preference matrix 1 ，U 2 )U 1 ！＝U 2 Obtaining a preference similarity matrix:
then, performing neighbor iteration recommendation by using a neighbor iteration unit 251C, specifically, sequencing the obtained similarity, setting a threshold h, and taking the similarity larger than h as a secondary neighbor; for the secondary neighbors not existing, calculating the similarity value of the secondary neighbors through the neighbor attenuation ratio and the attenuation weight; for the existence of the secondary neighbors, a larger value is taken as a new similarity by comparing the actual similarity with the similarity calculated by the attenuation weight, and then the candidate neighbors are ranked, and the first k are selected. All users after one selection form a new neighbor set, and iteration can be carried out on the basis of the new neighbor set to obtain a final neighbor set. The k value here is set according to the actual situation.
Then, the scoring prediction unit 251D is used for scoring prediction, the method uses the room source set I 'with the browsing duration less than the threshold value T for the user to calculate based on the new set U' of the previous k candidate neighbors and the associated users
Scoring (target user U) i For each room source I in I k Preference degree) Q i,k ：
V(Q i ) Representing the average preference value of user i
And finally, sorting the predicted scores obtained in the score prediction unit 251D according to the size, and taking the first 3 corresponding room sources as room source recommendation results.
In one embodiment, as shown in FIG. 4, the customer recommendation module 260 includes: a weight setting unit 261, a weight updating unit 262, a quality calculating unit 263, and a client recommending unit 264, wherein:
the weight setting unit 261 is configured to set weights W1 and W2 for the browsing duration T and the problem consultation of the client, and to sort questions and answers in the problem set according to the attention degree, and set different weights W;
the weight updating unit 262 is configured to update the attention degree of the problem in the problem set according to the client information of the client, and update the weight w;
the quality calculating unit 263 is used for calculating the quality of the client according to the weights W1, W2 and W;
the client recommendation unit 264 is configured to take a client with a quality greater than a threshold as a target client according to a preset threshold, and push the target client to the server through the communication module.
Specifically, firstly, designing weights W1 and W2 for browsing duration T and problem consultation of a client, sequencing a problem set according to the attention degree, and setting different weights W;
secondly, updating the attention degree of the problem in the problem set and updating the weight value w from the history dialogue of the client and the virtual broker;
then calculating the mass:wherein w is i The weight of the effective questions in the dialogue is that n represents the number of the effective questions;
and finally, designing a threshold, and when m is greater than or equal to the threshold, indicating that the quality of the client is good, and sending the client to a server for storage or pushing the client to customer service for follow-up of the client.
In one embodiment, the client recommendation unit 260 is further configured to push the target client to the knowledge base, and update the map. Specifically, the client recommendation unit 260 may push the target client into the knowledge base, corresponding to updating the portion of the graph with respect to the quality of the client.
In one embodiment, as shown in fig. 5 and 7, the intelligent answer module 270 includes: a question analysis sub-module 271, an information retrieval sub-module 272, and an answer generation sub-module 273, the question analysis sub-module 271 including an entity obtaining unit 271A and a destination obtaining unit 271B, wherein:
the entity obtaining unit 271A is configured to perform word segmentation and entity identification processing on a question of a client to obtain an entity of the question;
the objective obtaining unit 271B is configured to perform objective recognition and question classification processing on the client question, and obtain the objective of the question;
the information retrieval sub-module 272 is configured to query the knowledge base according to the entity and the purpose, and obtain an initial query result;
the answer generation sub-module 273 is configured to compare and rank the initial query results, select an optimal answer, generate a target answer in the form of a natural sentence from the optimal answer, and feed back the target answer to the client.
In particular, the questions are brought into a knowledge base-based question-answer that a knowledge base prepared in advance seeks answers. Analyzing a natural language question input by a user, inquiring a knowledge spectrogram database, finally obtaining an optimal answer, and then feeding back to the user in a natural language mode. Or after the voice is input, the voice is required to be converted into a text, then the questions are analyzed and the answers are predicted, and the predicted answers presented in natural language are obtained and then converted into the voice to be fed back to the user.
Firstly, word segmentation and entity recognition processing are carried out on a question of a client through an entity obtaining unit 271A, so that an entity of the question is obtained, specifically, a professional dictionary in the field (house property) needs to be established before word segmentation, then sentence segmentation is carried out by adopting a forward maximum matching algorithm in a word segmentation method based on the dictionary, finally, the unknown words in the dictionary are segmented by adopting an ICTCLAS word segmentation tool, parts of speech are marked, and named entities are recognized.
Then, the objective obtaining unit 271B performs objective recognition and question classification processing on the client questions, and obtains the objective of the questions, specifically, before objective recognition of the questions, it is necessary to build a domain-class corpus as an initial training set. The virtual broker is trained according to the training set to obtain a certain degree of identification capability. The training is mainly to extract characteristic values from a corpus, take the occurrence frequency of the segmented words as weight values, obtain a group of segmented words as characteristic vectors of a certain problem category according to the weight values, and take the weight values as the characteristic vector values. When a user presents a new question, the question is updated to the corpus. And simultaneously, predicting the problem to which the problem belongs to obtain a weight sequence, and taking the weight with the largest weight as the most probable purpose of the user.
After retrieving the entity and the destination by the information retrieving sub-module 272, the relationship between the entity and the destination is predicted by the entity obtaining unit 271A and the destination obtaining unit 271B, and then a triplet (Subject-Predict-Object) may be built to query the knowledge base to obtain the initial query result.
Finally, the initial query result is compared and sequenced through the answer generation sub-module 273, the best answer is selected, the best answer is generated into a target answer in the form of a natural sentence, and the target answer is fed back to the client.
In one embodiment, as shown in fig. 8, the intelligent answer module 270 further includes a natural language generation sub-module including a content determination unit 274A, a structure determination unit 274B, and a language generation unit 274C, wherein:
the content determining unit 274A is configured to use the entity and the target answer as sentence key information;
the structure determining unit 274B is configured to determine a sentence structure based on the question set and according to the purpose;
the language generating unit 274C is configured to combine the sentence key information in accordance with the sentence structure, thereby obtaining the target answer in the form of a natural sentence.
Specifically, first, the key content is determined: namely, the named entities identified in the question classification and answers obtained from the information retrieval are used as statement key information; secondly, determining a sentence structure: the method comprises the steps that the problems corresponding to the most probable intention of a user are obtained in the problem classification, and then sentence structure models are obtained from a corpus; finally, natural language realization is carried out: the phrases determined in the content determining unit 274A are combined into a well-formatted sentence in accordance with the model in the structure determining unit 274B.
It will be appreciated by those skilled in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device, or distributed across a network of computing devices, or they may alternatively be implemented in program code executable by computing devices, such that they may be stored on a computer storage medium (ROM/RAM, magnetic or optical disk) for execution by computing devices, and in some cases, the steps shown or described may be performed in a different order than what is shown or described herein, or they may be individually manufactured as individual integrated circuit modules, or a plurality of modules or steps in them may be manufactured as a single integrated circuit module. Therefore, the present invention is not limited to any specific combination of hardware and software.
The foregoing is a further detailed description of the invention in connection with specific embodiments, and is not intended to limit the practice of the invention to such descriptions. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.
Claims (7)
1. A virtual broker for use in house-seeing software, comprising: the system comprises an information receiving module, an information processing module, a storage module, a communication module, a house recommending module, a client recommending module and an intelligent answering module;
the system comprises a storage module, a database and a database, wherein the storage module is used for storing house source information, house introduction information and customer information, the database is used for storing a map and a problem set, the map is formed by integrating house source information, peripheral information of house source address positioning, a broker explanation and old customer evaluation and browsing records in the database, and the problem set is a historical question-answer set of a customer; the house source information at least comprises house source house types, house addresses and peripheral information; the house introduction information at least comprises house official explanation information and house official evaluation information;
the information processing module is used for integrating the information in the resource library into a map and a problem set in the knowledge base;
the communication module is communicated with the server, and high-quality clients are uploaded to the server through the communication module;
wherein, the customer recommendation module includes: a weight setting unit, a weight updating unit, a quality calculating unit and a client recommending unit, wherein:
the weight setting unit is used for setting weights W1 and W2 for browsing duration T and problem consultation of clients respectively, ordering questions and answers in the problem set according to the attention degree, and setting different weights W;
the weight updating unit is used for updating the attention degree of the problem in the problem set according to the client information of the client and updating the weight w;
the quality calculation unit is used for calculating the quality of the client according to the weights W1, W2 and W;
the client recommending unit is used for taking the client with the quality larger than the threshold value as a target client according to the preset threshold value and pushing the client to the server through the communication module;
the intelligent answer module can answer the customer questions based on the knowledge base.
2. A virtual broker for use in house-keeping software as recited in claim 1, wherein the customer information includes at least customer question-answer information, customer-browsing record information, and customer house rating information.
3. A virtual broker for use in house-keeping software as recited in claim 1, wherein the information receiving module is configured to receive a variety of information including house source information, house introduction information, and customer information.
4. A virtual broker for use in house-watching software as recited in claim 1, wherein the house recommendation module comprises: the system comprises a calculation and analysis submodule and a house recommendation submodule, wherein the calculation and analysis submodule comprises a model building unit, a similarity calculation unit, a neighbor iteration unit and a scoring prediction unit, wherein:
the model building unit is used for building a preference matrix model of a user about a house source by utilizing the information in the resource library;
the similarity calculation unit is used for calculating the similarity among the users by utilizing cosine similarity based on the preference matrix model to obtain a preference similarity matrix;
the neighbor iteration unit is used for obtaining new similarity among users according to the attenuation ratio and the attenuation weight based on the similarity of each user, and iterating the neighbor set according to a neighbor set formed by the new similarity to obtain a target neighbor set;
the scoring prediction unit is used for predicting the scoring of the user about the house source according to the client information based on the target neighbor set;
and the house recommending submodule is used for sorting the scores according to the size and selecting the top three house sources as target house sources to recommend to the customer.
5. A virtual broker for use in house-keeping software as recited in claim 1, wherein the client recommendation unit is further configured to push the target client into the knowledge base to update the profile.
6. A virtual broker for use in house-watching software as recited in claim 1, wherein the intelligent answer module comprises: the system comprises a question analysis sub-module, an information retrieval sub-module and an answer generation sub-module, wherein the question analysis sub-module comprises an entity obtaining unit and a purpose obtaining unit, and the question analysis sub-module comprises a target obtaining unit, wherein:
the entity obtaining unit is used for carrying out word segmentation and entity identification processing on the questioning of the client to obtain an entity of the questioning;
the objective obtaining unit is used for carrying out objective identification and problem classification processing on the client questions and obtaining the objectives of the questions;
the information retrieval submodule is used for inquiring the knowledge base according to the entity and the purpose and obtaining an initial inquiry result;
the answer generation submodule is used for comparing and sequencing the initial query results, selecting an optimal answer, generating a target answer in a natural sentence form from the optimal answer, and feeding the target answer back to the client.
7. A virtual broker for use in house-keeping software as recited in claim 6, wherein the intelligent answer module further comprises a natural language generation sub-module including a content determination unit, a structure determination unit, and a language generation unit, wherein:
the content determining unit is used for taking the entity and the target answer as statement key information;
the structure determining unit is used for determining a sentence structure based on the problem set and according to the purpose;
the language generating unit is used for combining the statement key information according to the statement structure so as to obtain a target answer in a natural statement form.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011050180.6A CN112330387B (en) | 2020-09-29 | 2020-09-29 | Virtual broker applied to house watching software |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011050180.6A CN112330387B (en) | 2020-09-29 | 2020-09-29 | Virtual broker applied to house watching software |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112330387A CN112330387A (en) | 2021-02-05 |
| CN112330387B true CN112330387B (en) | 2023-07-18 |
Family
ID=74313720
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011050180.6A Active CN112330387B (en) | 2020-09-29 | 2020-09-29 | Virtual broker applied to house watching software |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112330387B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113779210A (en) * | 2021-07-30 | 2021-12-10 | 深圳市思为软件技术有限公司 | Information introduction method, related device and computer readable storage medium |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010182287A (en) * | 2008-07-17 | 2010-08-19 | Steven C Kays | Intelligent adaptive design |
| CN103700003A (en) * | 2013-12-30 | 2014-04-02 | 陶鹏 | House online direct renting method and system based on wish conformity matching |
| CN104182919A (en) * | 2014-07-21 | 2014-12-03 | 安徽华贞信息科技有限公司 | Smart living informational service platform |
| CN205507888U (en) * | 2016-04-11 | 2016-08-24 | 东北石油大学 | Intelligent management system of house intermediary |
| CN108230051A (en) * | 2018-02-12 | 2018-06-29 | 昆山数泰数据技术有限公司 | A kind of user based on label Weight algorithm is to the determining method of commodity attention rate |
| CN108804529A (en) * | 2018-05-02 | 2018-11-13 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on Web |
| CN109658192A (en) * | 2018-12-20 | 2019-04-19 | 重庆锐云科技有限公司 | A kind of source of houses recommended method and server |
| CN109670113A (en) * | 2018-12-20 | 2019-04-23 | 重庆锐云科技有限公司 | A kind of source of houses recommended method, device and server |
| WO2019153613A1 (en) * | 2018-02-09 | 2019-08-15 | 平安科技（深圳）有限公司 | Chat response method, electronic device and storage medium |
| CN110298637A (en) * | 2019-07-04 | 2019-10-01 | 北京芒果铭天科技有限公司 | A kind of client management system |
| CN110298465A (en) * | 2019-07-04 | 2019-10-01 | 北京芒果铭天科技有限公司 | Houseclearing management system |
| CA3010392A1 (en) * | 2018-07-04 | 2020-01-04 | Quitchet, Llc | Social network application for real estate |
| CN110689457A (en) * | 2019-10-09 | 2020-01-14 | 重庆锐云科技有限公司 | Intelligent reception method for online clients in real estate industry, server and storage medium |
| CN110716979A (en) * | 2019-10-18 | 2020-01-21 | 重庆锐云科技有限公司 | House buying intention client mining method, device and server |
| CN110837604A (en) * | 2019-10-16 | 2020-02-25 | 贝壳技术有限公司 | Data analysis method and device based on housing monitoring platform |
| CN111639988A (en) * | 2020-05-22 | 2020-09-08 | 贝壳技术有限公司 | Broker recommendation method, device, electronic equipment and storage medium |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070050256A1 (en) * | 2000-06-30 | 2007-03-01 | Jay Walker | Method and apparatus for compensating participation in marketing research |
| US20020052814A1 (en) * | 2000-07-10 | 2002-05-02 | Ketterer Robert M. | Virtual real estate brokage system |
| US20030036944A1 (en) * | 2000-10-11 | 2003-02-20 | Lesandrini Jay William | Extensible business method with advertisement research as an example |
| US20030220807A1 (en) * | 2002-05-23 | 2003-11-27 | Kevin Hoffman | Automated method and system for managing and/or transferring real estate information |
| US10755369B2 (en) * | 2014-07-16 | 2020-08-25 | Parkland Center For Clinical Innovation | Client management tool system and method |
| US11093992B2 (en) * | 2017-05-05 | 2021-08-17 | Reai Inc. | Smart matching for real estate transactions |
| US11615463B2 (en) * | 2018-08-01 | 2023-03-28 | Appfolio, Inc. | Artificial intelligence based digital leasing assistant |
| US11423425B2 (en) * | 2019-01-24 | 2022-08-23 | Qualtrics, Llc | Digital survey creation by providing optimized suggested content |
-
2020
- 2020-09-29 CN CN202011050180.6A patent/CN112330387B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010182287A (en) * | 2008-07-17 | 2010-08-19 | Steven C Kays | Intelligent adaptive design |
| CN103700003A (en) * | 2013-12-30 | 2014-04-02 | 陶鹏 | House online direct renting method and system based on wish conformity matching |
| CN104182919A (en) * | 2014-07-21 | 2014-12-03 | 安徽华贞信息科技有限公司 | Smart living informational service platform |
| CN205507888U (en) * | 2016-04-11 | 2016-08-24 | 东北石油大学 | Intelligent management system of house intermediary |
| WO2019153613A1 (en) * | 2018-02-09 | 2019-08-15 | 平安科技（深圳）有限公司 | Chat response method, electronic device and storage medium |
| CN108230051A (en) * | 2018-02-12 | 2018-06-29 | 昆山数泰数据技术有限公司 | A kind of user based on label Weight algorithm is to the determining method of commodity attention rate |
| CN108804529A (en) * | 2018-05-02 | 2018-11-13 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on Web |
| CA3010392A1 (en) * | 2018-07-04 | 2020-01-04 | Quitchet, Llc | Social network application for real estate |
| CN109658192A (en) * | 2018-12-20 | 2019-04-19 | 重庆锐云科技有限公司 | A kind of source of houses recommended method and server |
| CN109670113A (en) * | 2018-12-20 | 2019-04-23 | 重庆锐云科技有限公司 | A kind of source of houses recommended method, device and server |
| CN110298637A (en) * | 2019-07-04 | 2019-10-01 | 北京芒果铭天科技有限公司 | A kind of client management system |
| CN110298465A (en) * | 2019-07-04 | 2019-10-01 | 北京芒果铭天科技有限公司 | Houseclearing management system |
| CN110689457A (en) * | 2019-10-09 | 2020-01-14 | 重庆锐云科技有限公司 | Intelligent reception method for online clients in real estate industry, server and storage medium |
| CN110837604A (en) * | 2019-10-16 | 2020-02-25 | 贝壳技术有限公司 | Data analysis method and device based on housing monitoring platform |
| CN110716979A (en) * | 2019-10-18 | 2020-01-21 | 重庆锐云科技有限公司 | House buying intention client mining method, device and server |
| CN111639988A (en) * | 2020-05-22 | 2020-09-08 | 贝壳技术有限公司 | Broker recommendation method, device, electronic equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 基于项目聚类和时间衰减的动态协同过滤算法;刘旭;李玲娟;;计算机技术与发展(第08期);全文 * |
| 融合社区划分的个性化美食推荐算法;陈诚;袁圆;余法红;蒋涛;周伟;潘海涛;滕泽伟;元丹;;福建电脑(第07期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112330387A (en) | 2021-02-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111444428B (en) | Information recommendation method and device based on artificial intelligence, electronic equipment and storage medium | |
| US20220215032A1 (en) | Ai-based recommendation method and apparatus, electronic device, and storage medium | |
| CN109284357B (en) | Man-machine conversation method, device, electronic equipment and computer readable medium | |
| CN108153876B (en) | Intelligent question and answer method and system | |
| WO2021159632A1 (en) | Intelligent questioning and answering method and apparatus, computer device, and computer storage medium | |
| CN109376222B (en) | Question-answer matching degree calculation method, question-answer automatic matching method and device | |
| JP5157314B2 (en) | Similarity calculation method, context model derivation method, similarity calculation program, context model derivation program | |
| CN117033608A (en) | Knowledge graph generation type question-answering method and system based on large language model | |
| CN112214670A (en) | Online course recommendation method and device, electronic equipment and storage medium | |
| CN110019736B (en) | Question-answer matching method, system, equipment and storage medium based on language model | |
| CN110674279A (en) | Question-answer processing method, device, equipment and storage medium based on artificial intelligence | |
| CN113505204B (en) | Recall model training method, search recall device and computer equipment | |
| CN110597962A (en) | Search result display method, device, medium and electronic equipment | |
| US11461613B2 (en) | Method and apparatus for multi-document question answering | |
| CN112905897B (en) | Similar user determination method, vector conversion model, device, medium and equipment | |
| CN112148831B (en) | Image-text mixed retrieval method and device, storage medium and computer equipment | |
| CN113239173A (en) | Method and device for processing question and answer data, storage medium and electronic equipment | |
| CN109977292B (en) | Search method, search device, computing equipment and computer-readable storage medium | |
| CN112330387B (en) | Virtual broker applied to house watching software | |
| CN114239730A (en) | Cross-modal retrieval method based on neighbor sorting relation | |
| CN116226320A (en) | Method and device for predicting context information, electronic equipment and storage medium | |
| CN116414940A (en) | Standard problem determining method and device and related equipment | |
| CN113094584A (en) | Method and device for determining recommended learning resources | |
| CN113342944B (en) | Corpus generalization method, apparatus, device and storage medium | |
| CN115129963A (en) | Search processing method and device |
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 |