CN105808688B - Complementary retrieval method and device based on artificial intelligence - Google Patents
Complementary retrieval method and device based on artificial intelligence Download PDFInfo
- Publication number
- CN105808688B CN105808688B CN201610119571.6A CN201610119571A CN105808688B CN 105808688 B CN105808688 B CN 105808688B CN 201610119571 A CN201610119571 A CN 201610119571A CN 105808688 B CN105808688 B CN 105808688B
- Authority
- CN
- China
- Prior art keywords
- query statement
- query
- user
- phrase
- word
- 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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application provides a completion retrieval method and a completion retrieval device based on artificial intelligence, wherein the method comprises the following steps: receiving a first query statement currently input by a user, and judging whether the first query statement meets completion conditions according to a preset completion retrieval strategy; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
Description
Technical Field
The application relates to the technical field of artificial intelligence information retrieval, in particular to a completion retrieval method and device based on artificial intelligence.
Background
Artificial Intelligence (Artificial Intelligence), abbreviated in english as AI. The method is a new technical science for researching and developing theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. Artificial intelligence is a branch of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a manner similar to human intelligence, a field of research that includes robotics, speech recognition, image recognition, natural language processing, and expert systems. Among them, the most important aspect of artificial intelligence is speech recognition technology.
With the popularization of the internet, the information on the internet is more and more abundant, and people can conveniently acquire the information wanted by themselves through a search engine.
In recent years, mobile internet using a smart phone as a terminal is rapidly expanded, and a large number of applications depending on the smart phone terminal are also produced. However, smart phones have special features compared to computers, such as: the screen size and keyboard input of the smart phone are much smaller than those of a computer; from the interactive time perspective, the smartphone user is more prone to fragmented interactions, such as gaps in public transportation. Therefore, the efficiency of search interaction by using the smart phone is low.
Disclosure of Invention
The present application is directed to solving, at least to some extent, one of the technical problems in the related art.
Therefore, a first objective of the present application is to provide a completion retrieval method based on artificial intelligence, which achieves to improve the efficiency of interaction between a search engine and a user through a multi-round retrieval manner, and to save the time for the user to input query statements.
A second object of the present application is to provide a supplementary retrieval device based on artificial intelligence.
In order to achieve the above object, an embodiment of a first aspect of the present application provides a completion retrieval method based on artificial intelligence, including: receiving a first query statement currently input by a user, and judging whether the first query statement meets completion conditions according to a preset completion retrieval strategy; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user.
According to the completion retrieval method based on artificial intelligence, whether a first query statement currently input by a user meets a completion condition is judged according to a preset completion retrieval strategy by receiving the first query statement; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
In order to achieve the above object, a second aspect of the present application provides an artificial intelligence-based completion retrieval apparatus, including: the receiving module is used for receiving a first query statement currently input by a user; the first judgment module is used for judging whether the first query statement meets completion conditions according to a preset completion retrieval strategy; a second judging module, configured to judge whether a second query statement retrieved by the user last time is related to the first query statement if the first query statement satisfies a completion condition; a completion module, configured to, if the second query statement is related to the first query statement, perform completion on the first query statement according to the second query statement to generate a third query statement; and the retrieval module is used for retrieving the third query statement and feeding back a retrieval result to the user.
According to the completion retrieval device based on artificial intelligence, whether a first query statement currently input by a user meets a completion condition or not is judged according to a preset completion retrieval strategy by receiving the first query statement; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
Drawings
The foregoing and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flow diagram of a method for artificial intelligence based completion retrieval in accordance with one embodiment of the present application;
FIG. 2 is a flow chart of a method of artificial intelligence based completion retrieval in accordance with another embodiment of the present application;
FIG. 3 is a technical block diagram of a multi-round completion system based on an entity attribute knowledge base;
FIG. 4 is a schematic diagram of a multi-round search by a completion retrieval approach;
FIG. 5 is an exemplary diagram of a conventional single-round search engine;
FIG. 6 is a schematic diagram of a search engine multi-round supplemental search;
fig. 7 is a schematic structural diagram of an artificial intelligence-based completion retrieval apparatus according to an embodiment of the present application.
Detailed Description
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
The following describes a complement retrieval method and apparatus based on artificial intelligence according to an embodiment of the present application with reference to the drawings.
Fig. 1 is a flowchart of an artificial intelligence based completion retrieval method according to an embodiment of the present application.
As shown in fig. 1, the artificial intelligence-based completion retrieval method includes:
Specifically, the completion retrieval method based on artificial intelligence provided by the embodiment of the invention is applied to terminal equipment with a retrieval function. Among them, the types of terminal devices are many, for example: smart phones, tablet computers, and the like.
When a first query statement currently input by a user is received, analyzing the first query statement according to a preset completion retrieval strategy, and judging whether the first query statement meets a completion condition. And if the first query statement does not meet the completion condition, directly retrieving the first query statement and feeding back a retrieval result to the user. And if the first query statement meets the completion condition, determining whether to perform completion retrieval on the first query statement according to the second query statement input last.
It should be noted that different completion retrieval strategies may be set according to different application needs, for example:
example one: the first query statement can be analyzed according to a preset phrase structure to judge whether the first query statement meets a completion condition; and/or the presence of a gas in the gas,
example two: whether the first query statement meets the completion condition can be judged according to whether the user inputs the query statement in a preset time period before the current time.
It should be noted that the above-mentioned completion retrieval strategy is only an example, and may be set or adjusted according to the actual application requirement.
Specifically, if a first query statement meets a preset completion condition, judging whether a second query statement retrieved by a user last time is related to the first query statement, and if the second query statement is related to the first query statement, performing completion retrieval on the first query statement; if the second query statement is not related to the first query statement, the first query statement is retrieved directly.
It should be noted that, different manners may be adopted according to different application needs, to determine whether the second query statement retrieved by the user last time is related to the first query statement retrieved currently, for example:
example one: the association degree of the second query statement with the first query statement may be evaluated according to a pre-trained discrimination model, and if the association degree satisfies a preset threshold, it is determined that the second query statement is related to the first query statement.
Example two: whether the second query statement is related to the first query statement can be judged according to the pre-stored phrase structure association information.
It should be noted that the above-mentioned manner for determining whether the first query statement and the second query statement are related is merely an example, and may be set or adjusted according to the actual application requirement.
And 104, retrieving the third query statement, and feeding back a retrieval result to the user.
Specifically, if it is determined that the second query statement retrieved last by the user is related to the currently retrieved first query statement, the currently retrieved first query statement is complemented according to the second query statement to generate a third query statement.
And then, searching the completed third query sentence, and feeding back a search result to the user.
According to the completion retrieval method based on artificial intelligence, whether a first query statement currently input by a user meets a completion condition is judged according to a preset completion retrieval strategy by receiving the first query statement; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
Fig. 2 is a flowchart of an artificial intelligence based completion retrieval method according to another embodiment of the present application.
As shown in fig. 2, the artificial intelligence based completion retrieval method may include the following steps:
Specifically, a phrase structure for detecting whether the first query statement satisfies the completion condition is preset, and a combination manner of the phrase structure may be set according to actual application needs, for example: the preset phrase structure comprises: the method comprises the steps that entity words and attribute words, when a first query statement is 'New York City captain', the query statement is known to comprise a preset phrase structure, wherein 'New York City' is an entity word; "city leader" is an attribute word. When the first query statement is "color", it is known that the query statement does not include a preset phrase structure.
If the first query statement does not contain a preset phrase structure, the first query statement meets a completion condition, and then step 203 is executed; if the first query statement contains a preset phrase structure, the first query statement does not satisfy a completion condition, and step 207 is executed;
Specifically, a time threshold is preset, and whether the user inputs a query statement within a preset time period from the current time to the time threshold is judged to determine whether a second query statement capable of completing the first query statement exists, so as to judge whether the first query statement meets a completion condition.
If the query sentence input by the user in the preset time period is judged and known, the first query sentence meets the completion condition, and then step 204 is executed; if it is determined that the user does not input the query statement within the preset time period, the first query statement does not satisfy the completion condition, and then step 207 is executed;
specifically, if the first query statement satisfies the completion condition, it is determined whether the first query statement is related to a second query statement detected by the user last time. First, whether the second query sentence contains a preset phrase structure is judged. The combination mode of the preset phrase structure can be set according to the actual application requirement.
If the second query statement contains the preset phrase structure, it is known that the second query statement is not related to the first query statement, and then step 207 is executed; if the second query statement does not contain the preset phrase structure, go to step 205;
specifically, a phrase relationship library is preset, wherein the relevance of each phrase structure in the phrase relationship library may be set according to actual application needs, for example: the phrase relation library comprises an entity word library, an attribute word library and a corresponding relation between each word in the entity word library and each word in the attribute word library.
If the phrase relationship library includes a first phrase in the first query statement and a second phrase in the second query statement, it indicates that the second query statement is related to the first query statement, and then step 206 is performed; if the phrase relationship library does not contain the first phrase in the first query statement and the second phrase in the second query statement, it indicates that the second query statement is not related to the first query statement, and then step 207 is executed.
And step 206, completing the first query statement according to the second query statement to generate a third query statement, retrieving the third query statement, and feeding back a retrieval result to the user.
Specifically, if it is determined that the second query statement retrieved last by the user is related to the currently retrieved first query statement, the currently retrieved first query statement is complemented according to the second query statement to generate a third query statement.
And then, searching the completed third query sentence, and feeding back a search result to the user.
And step 207, retrieving the first query statement, and feeding back a retrieval result to the user.
Specifically, if the first query statement does not satisfy the completion condition, or the first query statement and the second query statement do not satisfy the correlation, the first query statement is directly retrieved, and a retrieval result is fed back to the user.
According to the completion retrieval method based on artificial intelligence, whether a first query statement currently input by a user meets a completion condition is judged according to a preset completion retrieval strategy by receiving the first query statement; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
In order to more clearly illustrate the implementation process of the artificial intelligence-based completion retrieval method provided by the present application, the following embodiments are described in detail.
Fig. 3 is a technical block diagram of a multi-round completion system based on an entity attribute knowledge base. As shown in fig. 3, an important problem in the completion retrieval method based on artificial intelligence is how to judge that the query sentence input by the user needs to be subjected to completion retrieval. Firstly, judging whether the first query statement meets a completion condition according to a preset completion retrieval strategy; and if the first query statement meets a completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement.
How to judge whether the query statement input by the user is related to the previous query statement, namely whether the previous query statement needs to be used for completing the current query statement input by the user. For example, if the previous query statement is "san francisco majors", and the current query statement input by the user is "area", the "san francisco area" needs to be complemented, but the current query statement input by the user is "liu de hua", the completion is not needed, or the complemented query statement is itself, because the user opens a new search scene, "liu de hua", the query statement becomes the first query statement of the subsequent other query statements.
The strategy for judging whether completion is needed has high precision, and because completion errors can cause severe user experience, the method for judging whether completion is needed by using the entity attribute knowledge base is innovatively provided. The entity attribute knowledge base is a relational knowledge base of all entity words and attribute words, whether a given entity word (such as 'san francisco') and a given attribute word (such as 'city leader') have a relation or not can be judged according to the knowledge base, so that the entity word and the attribute word of a last query sentence and a current sentence input by a user need to be recognized respectively when completing the query, a forward maximum matching method in the existing word segmentation technology is used in the recognition method, and if the last query sentence or the current query sentence has other words besides the entity word and the attribute word, the solution is out of the solution range of the method, namely the current query sentence is not completed. After the entity words and the attribute words are identified, if the current sentence has the entity words and the attribute words, the current sentence is considered to have complete components without completion. When the current sentence only has the entity word A and the previous sentence has the attribute word B, judging whether the A and the B have a relationship, if so, completing the entity word of the current sentence by using the attribute word of the previous sentence; when the current sentence only has the attribute word B and the previous sentence has the entity word A, judging whether the A and the B have a relationship, if so, completing the attribute word of the current sentence by using the entity word of the previous sentence.
Fig. 4 is a schematic diagram of implementing multiple rounds of search by means of completion retrieval. Referring to fig. 4, an important problem in multiple rounds of sessions is how a search engine adds information of a previous query statement, or even multiple query statements, to a current query after a user inputs a query statement, and the implementation manner of this document is to complement the query statement input by the user with the information of the previous query statement into a complete query statement, and the search engine only uses the complemented query statement to perform retrieval without combining the previous query statement, or even the previous query statements to perform analysis, i.e., the complemented query statement and a conventional single round of search engine can implement multiple rounds of search engines, and a core retrieval system and a retrieval policy of the search engine do not need to be changed.
It should be emphasized that the previous original query statement, but the previous complemented query statement is used to complement the query statement input by the user. As shown in fig. 5, when the third query statement input by the user is complemented, the second original query statement "majordomo" is not used, but the second complemented query statement "majordomo" is used, and similarly, if there is a fourth query statement, the third complemented query statement is used for the alignment complementation, so that the recursive complementation can conveniently and continuously transfer multiple rounds of scenes "majordomo" from the first query statement to all subsequent query statements, and the first or second query statement is not used when the query statement input by the user is complemented.
Judging whether IP initiating the query statement or a user initiates other query statements within the first few minutes (generally 10 minutes) of initiating the query statement, if not, the query statement is the first query statement, and the sentence completed by the first query statement is forcibly set as the first query statement.
Fig. 5 is a diagram of an example of a conventional one-round search engine. Referring to fig. 5, a user inputs a query sentence, a search engine provides a query result, the user inputs another query sentence, the search engine provides a new query result, and the two query sentences input before and after the user appear to be irrelevant to the search engine, i.e., the first query result is only relevant to the first query sentence, and the second query result is only relevant to the second query sentence.
Fig. 6 is a schematic diagram of a search engine multi-round supplemental search. Referring to fig. 6, when a user inputs a first query statement, a search engine provides a query result, when the user inputs a second query statement, the search engine provides the query result according to the first query statement and in combination with the previous query statement or even the previous query statements, the multi-round conversational search interaction manner really realizes scene interaction, interaction is performed around a scene of "san francisco" in a city in a schematic diagram, the first query statement provides an interaction scene, subsequent query statements do not need to repeatedly describe the scene, only the query content in the scene needs to be indicated, that is, when the user needs to query "san francisco mao", only the "mao" needs to be input, and 60% of input time of the user can be saved.
In order to realize the embodiment, the application also provides a completion retrieval device based on artificial intelligence.
Fig. 7 is a schematic structural diagram of an artificial intelligence-based completion retrieval apparatus according to an embodiment of the present application.
As shown in fig. 7, the artificial intelligence-based completion retrieval apparatus includes:
a receiving module 11, configured to receive a first query statement currently input by a user;
a first judging module 12, configured to judge whether the first query statement satisfies a completion condition according to a preset completion retrieval policy;
a second judging module 13, configured to judge whether a second query statement retrieved by the user last time is related to the first query statement if the first query statement satisfies a completion condition;
a completion module 14, configured to, if the second query statement is related to the first query statement, perform completion on the first query statement according to the second query statement to generate a third query statement;
and the retrieval module 15 is configured to retrieve the third query statement and feed back a retrieval result to the user.
It should be noted that the foregoing explanation of the embodiment of the artificial intelligence based completion retrieval method is also applicable to the artificial intelligence based completion retrieval apparatus of this embodiment, and is not repeated here.
According to the completion retrieval device based on artificial intelligence, whether a first query statement currently input by a user meets a completion condition or not is judged according to a preset completion retrieval strategy by receiving the first query statement; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
Based on the above embodiment, further, the retrieving module 15 is further configured to:
and if the first query statement does not meet the completion condition, retrieving the first query statement and feeding back a retrieval result to the user.
Further, the retrieving module 15 is further configured to:
and if the second query statement is not related to the first query statement, retrieving the first query statement, and feeding back a retrieval result to the user.
In one embodiment, the first determining module 12 is configured to:
judging whether the first query sentence contains a preset phrase structure;
if the first query statement does not contain a preset phrase structure, judging whether the user inputs the query statement or not within a preset time period before the current moment;
and if the user inputs the query statement in the preset time period is judged and known, the first query statement meets a completion condition.
Wherein, the preset phrase structure comprises:
entity words and attribute words.
Further, the first determining module 12 is further configured to:
and if the first query statement contains a preset phrase structure, the first query statement does not meet a completion condition.
Further, the first determining module 12 is further configured to:
and if judging that the user does not input the query statement in the preset time period, the first query statement does not meet the completion condition.
In one embodiment, the second determining module 13 is configured to:
judging whether the second query sentence contains a preset phrase structure;
and if the second query statement contains a preset phrase structure, the second query statement is irrelevant to the first query statement.
Further, the second determining module 13 is further configured to:
if the second query statement does not contain a preset phrase structure, judging whether a first phrase in the first query statement and a second phrase in the second query statement have correlation or not according to a preset phrase relation library;
if the first phrase and the second phrase have relevance, the second query statement is relevant to the first query statement;
the second query statement is not related to the first query statement if the first phrase and the second phrase do not have a correlation.
Wherein, the preset phrase relation library comprises:
an entity word bank and an attribute word bank, and the corresponding relation between each word of the entity word bank and each word of the attribute word bank.
According to the completion retrieval device based on artificial intelligence, whether a first query statement currently input by a user meets a completion condition or not is judged according to a preset completion retrieval strategy by receiving the first query statement; if the first query statement meets the completion condition, judging whether a second query statement retrieved last by the user is related to the first query statement; if the second query statement is related to the first query statement, completing the first query statement according to the second query statement to generate a third query statement; and searching the third query statement, and feeding back a search result to the user. Therefore, the efficiency of interaction between the search engine and the user is improved in a multi-round retrieval mode, and the time for the user to input the query sentence is saved.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.
Claims (14)
1. A completion retrieval method based on artificial intelligence is characterized by comprising the following steps:
receiving a first query statement currently input by a user, and judging whether the first query statement meets completion conditions according to a preset completion retrieval strategy, wherein the judging whether the first query statement meets the completion conditions according to the preset completion retrieval strategy comprises the following steps: judging whether the first query sentence contains a phrase structure comprising entity words and attribute words, and if the first query sentence does not contain the complete phrase structure, judging whether the user inputs the query sentence in a preset time period before the current time;
if it is determined that the first query statement meets completion conditions after the query statement is input by the user within the preset time period, determining whether a second query statement retrieved last by the user is related to the first query statement, wherein when the first query statement only contains entity words and the second query statement includes attribute words, or when the first query statement only contains attribute words and the second query statement includes entity words, determining whether the second query statement retrieved last by the user is related to the first query statement, includes: judging whether the second query sentence searched last contains at least part of phrase structures lacked in the first query sentence, if so, judging whether the phrase structures contained in the second query sentence are related to at least part of phrase structures contained in the first query sentence;
if the second query statement is related to the first query statement, judging whether the entity word is related to the attribute word or not when the first query statement only contains the entity word and the second query statement comprises the attribute word, completing the entity word in the first query statement according to the attribute word if the entity word is related to the attribute word to generate a third query statement, judging whether the entity word is related to the attribute word or not when the first query statement only contains the attribute word and the second query statement comprises the entity word, completing the attribute word in the first query statement according to the entity word if the entity word is related to the attribute word to generate the third query statement;
and searching the third query statement, and feeding back a search result to the user.
2. The method of claim 1, further comprising:
and if the first query statement does not meet the completion condition, retrieving the first query statement and feeding back a retrieval result to the user.
3. The method of claim 1, further comprising:
and if the second query statement is not related to the first query statement, retrieving the first query statement, and feeding back a retrieval result to the user.
4. The method of claim 1, wherein the determining whether the first query statement satisfies a completion condition according to a preset completion retrieval policy further comprises:
and if the first query statement contains a preset phrase structure, the first query statement does not meet a completion condition.
5. The method of claim 1, wherein the determining whether the first query statement satisfies a completion condition according to a preset completion retrieval policy further comprises:
and if judging that the user does not input the query statement in the preset time period, the first query statement does not meet the completion condition.
6. The method of claim 1, wherein said determining whether a second query statement last retrieved by the user is related to the first query statement, further comprises:
if the second query statement does not contain a preset phrase structure, judging whether a first phrase in the first query statement and a second phrase in the second query statement have correlation or not according to a preset phrase relation library;
if the first phrase and the second phrase have relevance, the second query statement is relevant to the first query statement;
the second query statement is not related to the first query statement if the first phrase and the second phrase do not have a correlation.
7. The method of claim 6, wherein the predetermined phrase relationship library comprises:
an entity word bank and an attribute word bank, and the corresponding relation between each word of the entity word bank and each word of the attribute word bank.
8. The utility model provides a completion retrieval device based on artificial intelligence which characterized in that includes:
the receiving module is used for receiving a first query statement currently input by a user;
the first judging module is used for judging whether the first query sentence contains a phrase structure comprising entity words and attribute words or not, and judging whether the user inputs the query sentence in a preset time period before the current moment or not when the first query sentence does not contain the complete phrase structure;
a second determining module, configured to determine whether a second query statement retrieved by a previous user is related to a first query statement if it is determined that the first query statement satisfies a completion condition when it is determined that the user inputs the query statement within the preset time period, where determining whether the second query statement retrieved by the previous user is related to the first query statement if the first query statement only includes an entity word and the second query statement includes an attribute word, or if the first query statement only includes an attribute word and the second query statement includes an entity word, the determining whether the second query statement retrieved by the previous user is related to the first query statement includes: judging whether the second query sentence searched last contains at least part of phrase structures lacked in the first query sentence, if so, judging whether the phrase structures contained in the second query sentence are related to at least part of phrase structures contained in the first query sentence;
a completion module, configured to determine whether the entity word and the attribute word are related when the first query statement only includes an entity word and the second query statement includes an attribute word if the first query statement is related to the first query statement, perform completion on the entity word in the first query statement according to the attribute word if the first query statement includes the attribute word to generate a third query statement, determine whether the entity word and the attribute word are related when the first query statement only includes the attribute word and the second query statement includes the entity word, and perform completion on the attribute word in the first query statement according to the entity word if the first query statement includes the attribute word to generate a third query statement;
and the retrieval module is used for retrieving the third query statement and feeding back a retrieval result to the user.
9. The apparatus of claim 8, wherein the retrieval module is further to:
and if the first query statement does not meet the completion condition, retrieving the first query statement and feeding back a retrieval result to the user.
10. The apparatus of claim 9, wherein the retrieval module is further to:
and if the second query statement is not related to the first query statement, retrieving the first query statement, and feeding back a retrieval result to the user.
11. The apparatus of claim 8, wherein the first determining module is further configured to:
and if the first query statement contains a preset phrase structure, the first query statement does not meet a completion condition.
12. The apparatus of claim 8, wherein the first determining module is further configured to:
and if judging that the user does not input the query statement in the preset time period, the first query statement does not meet the completion condition.
13. The apparatus of claim 8, wherein the second determining module is further configured to:
if the second query statement does not contain a preset phrase structure, judging whether a first phrase in the first query statement and a second phrase in the second query statement have correlation or not according to a preset phrase relation library;
if the first phrase and the second phrase have relevance, the second query statement is relevant to the first query statement;
the second query statement is not related to the first query statement if the first phrase and the second phrase do not have a correlation.
14. The apparatus of claim 13, wherein the predetermined phrase relationship library comprises:
an entity word bank and an attribute word bank, and the corresponding relation between each word of the entity word bank and each word of the attribute word bank.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610119571.6A CN105808688B (en) | 2016-03-02 | 2016-03-02 | Complementary retrieval method and device based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610119571.6A CN105808688B (en) | 2016-03-02 | 2016-03-02 | Complementary retrieval method and device based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105808688A CN105808688A (en) | 2016-07-27 |
| CN105808688B true CN105808688B (en) | 2021-02-05 |
Family
ID=56466590
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610119571.6A Active CN105808688B (en) | 2016-03-02 | 2016-03-02 | Complementary retrieval method and device based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105808688B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106991152B (en) * | 2017-03-28 | 2020-12-11 | 百度在线网络技术（北京）有限公司 | Artificial intelligence-based on-demand resource retrieval method and device and terminal equipment |
| CN108563637A (en) * | 2018-04-13 | 2018-09-21 | 北京理工大学 | A kind of sentence entity complementing method of fusion triple knowledge base |
| CN109857845B (en) * | 2019-01-03 | 2021-06-22 | 北京奇艺世纪科技有限公司 | Model training and data retrieval method, device, terminal and computer-readable storage medium |
| CN114090722B (en) * | 2022-01-19 | 2022-04-22 | 支付宝(杭州)信息技术有限公司 | Method and device for automatically completing query content |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20050005214A (en) * | 2003-07-07 | 2005-01-13 | 정원철 | lnteractive Search Method |
| US9639618B2 (en) * | 2006-05-25 | 2017-05-02 | International Business Machines Corporation | System, method and program product for key word searching where a key word is a homonym |
| CN103678358A (en) * | 2012-09-13 | 2014-03-26 | 腾讯科技（深圳）有限公司 | Information search method and system |
| CN105912669B (en) * | 2012-09-20 | 2020-04-07 | 北京奇付通科技有限公司 | Method and device for complementing search terms and establishing individual interest model |
| CN103383701A (en) * | 2013-07-12 | 2013-11-06 | 北京小米科技有限责任公司 | Information retrieving method, device and terminal |
| CN104462505A (en) * | 2014-12-19 | 2015-03-25 | 北京奇虎科技有限公司 | Search method and device |
-
2016
- 2016-03-02 CN CN201610119571.6A patent/CN105808688B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN105808688A (en) | 2016-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102915299B (en) | Word segmentation method and device | |
| CN105808688B (en) | Complementary retrieval method and device based on artificial intelligence | |
| CN109657054A (en) | Abstraction generating method, device, server and storage medium | |
| CN109977215B (en) | Statement recommendation method and device based on associated interest points | |
| CN103488796B (en) | Based on context the method and mobile terminal inputted | |
| WO2013170587A1 (en) | Multimedia question and answer system and method | |
| CN110795553B (en) | Digest generation method and device | |
| CN106098063A (en) | A kind of sound control method, terminal unit and server | |
| US11361759B2 (en) | Methods and systems for automatic generation and convergence of keywords and/or keyphrases from a media | |
| CN109710732B (en) | Information query method, device, storage medium and electronic equipment | |
| CN106980652B (en) | Intelligent question and answer method and system | |
| US9811515B2 (en) | Annotating posts in a forum thread with improved data | |
| CN105893351B (en) | Audio recognition method and device | |
| CN112487810A (en) | Intelligent customer service method, device, equipment and storage medium | |
| US20160335267A1 (en) | Method and apparatus for natural language search for variables | |
| CN110442855A (en) | A kind of speech analysis method and system | |
| CN114995729A (en) | Voice drawing method and device and computer equipment | |
| CN106653006B (en) | Searching method and device based on interactive voice | |
| CN109063127A (en) | A kind of searching method, device, server and storage medium | |
| CN117648422A (en) | Question-answer prompt system, question-answer prompt, library construction and model training method and device | |
| WO2021103594A1 (en) | Tacitness degree detection method and device, server and readable storage medium | |
| CN117093600A (en) | Search prompt word generation method and device, electronic equipment and storage medium | |
| EP4322066A1 (en) | Method and apparatus for generating training data | |
| CN107918606B (en) | Method and device for identifying avatar nouns and computer readable storage medium | |
| US20220262353A1 (en) | Method and device for Processing Voice Information, Storage Medium and Electronic Apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |