CN109710935B - Museum navigation and knowledge recommendation method based on cultural relic knowledge graph - Google Patents
Museum navigation and knowledge recommendation method based on cultural relic knowledge graph Download PDFInfo
- Publication number
- CN109710935B CN109710935B CN201811604829.7A CN201811604829A CN109710935B CN 109710935 B CN109710935 B CN 109710935B CN 201811604829 A CN201811604829 A CN 201811604829A CN 109710935 B CN109710935 B CN 109710935B
- Authority
- CN
- China
- Prior art keywords
- knowledge
- cultural relic
- cultural
- query
- user
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a museum cultural relic navigation and knowledge recommendation method based on a knowledge graph, which comprises the steps of firstly converting natural language query input by a user into SPARQL query sentences by utilizing a word segmentation tool and a regular expression conversion rule, carrying out cultural relic knowledge query and reasoning in a SPARQL language mode, and visually presenting queried cultural relics to the user in a knowledge graph mode; meanwhile, the cultural relic or cultural relic background knowledge related to the current inquired cultural relic is recommended to the user through recommendation algorithms such as cultural relic attribute association and collaborative filtering, so that the interested user can select and interactively browse in a knowledge graph mode. The invention can interact with the user through natural language based on the established cultural relic knowledge graph, recommend other cultural relics related to the current cultural relics or possibly interested by the user to the user, and guide the user to browse the cultural relic information and know the knowledge of the related cultural relics better.
Description
Technical Field
The invention belongs to the field of knowledge maps, and particularly relates to a museum navigation and knowledge recommendation method based on a cultural relic knowledge map.
Background
In recent years, with the improvement of living standard of people, more and more people select tourism to increase knowledge and cultivate sentiment, and the museum is more and more popular among tourists as a elegant cultural tourism place. However, since there are many cultural relics exhibited in a museum, a visitor cannot directly obtain information of the cultural relics of interest, a method for applying a knowledge map to a museum navigation has been developed, and the visitor can be further helped to view the cultural relics from the museum.
However, the navigation method and the recommendation method in the prior art still have a big problem, wherein the navigation method cannot identify and query the natural language of people, which is far from the requirement of the intelligent query in the current era, and in the recommendation method, the content association algorithm cannot be adjusted according to the use preference of the user, and the collaborative filtering algorithm needs a large amount of cold start data when the use frequency of the user is small, which easily causes data loss and causes unnecessary problems.
Therefore, on the basis of establishing a good knowledge graph, a person skilled in the art needs to solve the problems that how to improve a query mode to support query by using natural language and how to solve the problems that the prior recommendation method is cold-started and cannot be adjusted according to preferences to improve the use and viewing experience of tourists.
Disclosure of Invention
In view of the above, the invention provides a museum navigation and knowledge recommendation method based on a cultural relic knowledge graph, which interacts with the knowledge graph through natural language processing and SPARQL query according to the currently generated knowledge graph, and recommends related cultural relics and related cultural relic knowledge to a user through a recommendation algorithm, thereby achieving the purpose of intelligent navigation.
In order to achieve the purpose, the invention adopts the following technical scheme:
a museum navigation and knowledge recommendation method based on cultural relic knowledge graph comprises the following steps:
the method comprises the following steps: acquiring user input;
step two: designing a word segmentation tool aiming at cultural relic knowledge;
step three: the user input is normalized by entity resolving entity associations and then participled by a participle tool.
Step four: completing the query by executing an SPARQL statement on the word segmentation result obtained in the step three;
step five: generating an association relation table by adopting a recommendation algorithm, and recommending related cultural relics in sequence according to the strength of association;
step six: and sorting the query result returned by the SPARQL and the recommendation result returned by the recommendation algorithm, and visually displaying.
Preferably, the first step specifically includes:
the user input can be character input or voice input;
if the input is voice input, the input needs to be converted into characters by calling a voice character conversion tool.
Preferably, the word segmentation tool for the cultural relic knowledge in the second step is implemented by adding proper nouns contained in the attribute of the cultural relic to a user dictionary of the word segmentation tool.
Preferably, the third step specifically includes the following steps:
the query statement is normalized by disambiguating the user input through an entity resolution and entity association method, and is divided and part-of-speech labeled by using a word segmentation tool.
Preferably, the step four specifically includes the following steps:
step 1): analyzing all possible word segmentation results in advance, sorting the word segmentation results into corresponding rules, and setting SPARQL query sentences corresponding to the rules;
step 2): and matching the obtained word segmentation result with the rule, and successfully executing the corresponding query statement.
Preferably, the recommendation algorithm in the fifth step includes a content association algorithm and a collaborative filtering algorithm;
firstly, setting a query threshold value as S, and adopting the content association algorithm when query records of all users are smaller than the query threshold value; and gradually transitioning the content association algorithm into a collaborative filtering algorithm along with the increase of the query records.
It should be noted that, the specific steps of the content association algorithm are as follows:
step 1): constructing a weight value table, and setting corresponding weight values for different attributes;
step 2): establishing an association relation table with an initial value of 0; traversing all the cultural relics in the incidence relation table, if the cultural relics in the incidence relation table are the same as a certain attribute of the cultural relics in the current query result, adding the weight values of the corresponding attributes in the incidence relation table according to the weight value table, and calculating a total value;
step 3): sorting the total values of the Chinese matters in the association relation table; the larger the total value is, the greater the correlation between the cultural relics and the cultural relics in the current query result is; and constructing a recommendation table containing n elements, and filling the first n cultural relics with the maximum correlation in the association relation table into the recommendation table.
Preferably, the specific step of gradually transitioning the content association algorithm into the collaborative filtering algorithm is as follows:
every time when the query record is increased by one S, subtracting 1 from the recommended number of the cultural relics obtained by the content association algorithm in the recommendation list, and adding 1 to the recommended number of the cultural relics obtained by the collaborative filtering algorithm until the recommended cultural relic results are all from the collaborative recommendation algorithm or the recommended cultural relic number obtained by the content association algorithm is 0.
It should be noted that, the specific steps of the collaborative filtering algorithm are as follows:
first, the behavior of the user is obtained by collecting the usage data of the user. The representation can be performed by an m x n order matrix. Element a of the matrixijIndicating that a user i has feedback on the recommendation j.
And then, a user set similar to the historical behavior data of the user is calculated, and the user set is called as a user neighbor.
And finally, finding out the most possible N objects from the recommended items selected by the similar user set to recommend the target user.
Through the technical scheme, compared with the prior art, the invention discloses a museum navigation and knowledge recommendation method based on the cultural relic knowledge graph.
Firstly, the invention realizes simple understanding of natural language by using word segmentation tool and regular expression, so that the user can complete query operation through natural language instead of according to a certain predetermined format, thereby saving learning time of the user.
In addition, the invention realizes the function of recommending the relevant cultural relics to the user by adding a recommendation algorithm. And the recommendation algorithm is an algorithm combining a content association algorithm and a collaborative filtering algorithm, so that the problem of cold start of a program when the use records are few is solved, and cultural relics which are more in line with the preference of the user can be recommended to the user when the use records are more.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flow chart of a museum navigation and knowledge recommendation method based on cultural relic knowledge graph according to the present invention;
FIG. 2 is a flow chart of a recommendation algorithm provided by the present invention;
FIG. 3 is a diagram illustrating the effect of query results and related recommendations on a visual interface according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to meet the requirements of querying museum cultural relic information by using a natural language and making recommendation of related cultural relics for a user by means of a recommendation algorithm, the embodiment of the invention provides a museum navigation and knowledge recommendation method based on a cultural relic knowledge graph.
As shown in fig. 1, the method mainly comprises two parts: inquiring historical relic knowledge and recommending related historical relics. Firstly, a word segmentation tool and a regular expression conversion rule are utilized to convert natural language query input by a user into SPARQL query sentences, cultural relic knowledge query and reasoning are carried out in a SPARQL language mode, and queried cultural relics are fed back to a GUI interface in a knowledge map mode to be presented to the user visually; meanwhile, the cultural relic or cultural relic background knowledge related to the current inquired cultural relic is recommended to the user through recommendation algorithms such as cultural relic attribute association and collaborative filtering, so that the interested user can select and interactively browse in a knowledge graph mode.
The method comprises the following steps of:
the method comprises the following steps: processing a query input by a user;
the form of the user-entered query is not specified herein, including but not limited to text, speech. If the input content is in the form of characters, the input content can be directly handed to the next step for processing. If the input content is in a voice form, the voice is converted into characters through a voice recognition tool, and the obtained characters are handed to the next step for processing.
Step two: designing a word segmentation tool aiming at cultural relic knowledge;
the method is characterized in that proper nouns related to cultural relic knowledge, such as cultural relic names, cultural relic attribute values and the like, are added into a user dictionary of the existing word segmentation tool through processing the cultural relic structured data, so that the new word segmentation tool can successfully complete word segmentation when the proper nouns related to the cultural relics are involved, wherein the proper nouns comprise related vocabularies such as heading names, cultural relic names, land names and use names.
Step three: the user input is normalized by entity resolving entity associations and then participled by a participle tool.
When a query is input, a plurality of queries may be input for the same query job, for example, when inquiring the land where a certain cultural relic is going out, a question may be directly asked about "where the place where a certain cultural relic is going out of the land", a question may also be asked about "where a certain cultural relic is going out of the land", and even a question may be asked about "is a certain cultural relic out of the land? These three problems are essentially the same problem. Here, we perform entity resolution and entity association on the input, and associate "where the earth is," where the earth is "and" is the earth is "to the same relationship entity, namely" earth is "of the relationship entity.
Step four: aiming at the word segmentation result obtained in the step three, executing a corresponding SPARQL statement according to a rule to perform query;
furthermore, when the input natural language is processed, two methods, namely rule matching and lexical analysis, are mainly used. For natural language with clear target, such as query cultural relics, high-frequency keywords such as "dynasty" and "province" may often appear, several sentence rules can be constructed by using the high-frequency keywords, and when the established rules are matched by using the regular expressions, the query result is returned according to the corresponding query sentence.
If the regular expression fails, the word segmentation tool designed for the cultural relic knowledge in the step two needs to be used for processing, after the normalization of the step three, the natural language input often relates to certain entity names of concepts such as dynasty, cultural relic names, land appearing and purposes, and if the keywords can be found in the input query sentence, the query intention of the user can be easily judged. And analyzing sentence patterns easily related to place nouns, time nouns, relation words, connection words and the like in the sentences according to the word segmentation results, setting corresponding models of word segmentation structures, executing corresponding query sentences when matching the input sentences, and returning query results.
The recommendation steps of the related cultural relics are as follows:
step five: constructing a recommendation algorithm, generating an association relation table, and recommending related cultural relics in sequence according to the strength of association; the recommendation algorithm comprises a content association algorithm and a collaborative filtering algorithm; firstly, setting a query record threshold value, and adopting a content association algorithm when the use record is smaller than the query record threshold value; when the usage record is greater than the query record threshold, a collaborative filtering algorithm is adopted, and the implementation flow is shown in fig. 2.
When the usage record is less, a threshold value is preset to measure, for example 2000, and when the usage record is less than the threshold value, the content association algorithm is completely adopted. Corresponding weight values are set for different attributes, and the parenthesis is the setting of one weight value corresponding to the attribute: dynasty (1), province (1), application (2) and detail dynasty (3).
And establishing a list of all the cultural relics related to the current cultural relic, wherein the initial value is 0. When the attribute of a cultural relic in the history list is the same as one or more attributes of the current cultural relic, the corresponding weight value is added to the attribute in the list. And finally, after all traversal is finished, sorting the total value of the attribute weights of the cultural relics in the list from large to small, selecting the first n cultural relics as recommended cultural relics to recommend to a user, and after a threshold value is reached, reducing the recommended quantity of the cultural relics from content association in the related recommendation by 1 and increasing the recommended quantity of the cultural relics from collaborative filtering by 1 every time when 2000 pieces of usage records are added until all the related recommendations come from a collaborative filtering algorithm, thereby ensuring that different usage records can generate different influences on the result under the condition that the quantity of the recommended cultural relics is not changed, and simultaneously solving the problem of cold start.
Step six: the query result returned by the SPARQL and the recommendation result returned by the recommendation algorithm are sorted and transmitted to an output interface of a visual interface, and the output interface is sorted through the visual interface and then presented to a user, wherein the presentation effect is shown in FIG. 3.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (5)
1. A museum navigation and knowledge recommendation method based on cultural relic knowledge graph is characterized by comprising the following steps:
the method comprises the following steps: acquiring user input;
step two: designing a word segmentation tool aiming at cultural relic knowledge;
step three: normalizing the user input through entity resolution and entity association, and performing word segmentation on the input by using a word segmentation tool;
step four: completing the query by executing an SPARQL statement on the word segmentation result obtained in the step three;
step five: generating an association relation table by adopting a recommendation algorithm, and recommending related cultural relics in sequence according to the strength of association;
step six: sorting the query result returned by the SPARQL and the recommendation result returned by the recommendation algorithm, and visually displaying the results;
the recommendation algorithm in the fifth step comprises a content association algorithm and a collaborative filtering algorithm;
firstly, setting a query threshold value as S, and adopting the content association algorithm when query records of all users are smaller than S; gradually transitioning the content association algorithm to the collaborative filtering algorithm as the query record increases;
the specific steps of gradually transitioning the content association algorithm to the collaborative filtering algorithm are as follows:
every time the query record is increased by one S, subtracting 1 from the recommended number of the cultural relics obtained by the content association algorithm in the recommendation list, and adding 1 to the recommended number of the cultural relics obtained by the collaborative filtering algorithm until the recommended cultural relic results are all from the collaborative recommendation algorithm or the recommended cultural relic number obtained by the content association algorithm is 0.
2. The cultural relic knowledge graph-based museum navigation and knowledge recommendation method according to claim 1, wherein the first step specifically comprises:
the user input comprises text input and voice input;
the voice input is converted into characters by calling a voice-to-character tool.
3. The cultural relic knowledge graph-based museum navigation and knowledge recommendation method according to claim 1, wherein the second step specifically comprises the following steps:
adding proper nouns containing the attributes of the cultural relics to a user dictionary of the word segmentation tool.
4. The cultural relic knowledge graph-based museum navigation and knowledge recommendation method according to claim 1, wherein the third step specifically comprises the following contents:
disambiguating the user input through an entity resolution and entity association method so as to standardize the query statement, and dividing and labeling the query statement by using the word segmentation tool.
5. The cultural relic knowledge graph-based museum navigation and knowledge recommendation method according to claim 1, wherein the fourth step specifically comprises the following steps:
step 1): analyzing all possible word segmentation results in advance, sorting the word segmentation results into corresponding rules, and setting SPARQL query sentences corresponding to the rules;
step 2): and matching the word segmentation result obtained after the word segmentation is carried out by the user input with the rule, and executing the corresponding query sentence if the matching is successful.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811604829.7A CN109710935B (en) | 2018-12-26 | 2018-12-26 | Museum navigation and knowledge recommendation method based on cultural relic knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811604829.7A CN109710935B (en) | 2018-12-26 | 2018-12-26 | Museum navigation and knowledge recommendation method based on cultural relic knowledge graph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109710935A CN109710935A (en) | 2019-05-03 |
| CN109710935B true CN109710935B (en) | 2021-03-26 |
Family
ID=66257742
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811604829.7A Active CN109710935B (en) | 2018-12-26 | 2018-12-26 | Museum navigation and knowledge recommendation method based on cultural relic knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109710935B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110428814B (en) * | 2019-07-25 | 2022-03-01 | 杭州蓦然认知科技有限公司 | Voice recognition method and device |
| CN110704411B (en) * | 2019-09-27 | 2022-12-09 | 京东方科技集团股份有限公司 | Knowledge graph building method and device suitable for art field and electronic equipment |
| CN111339320B (en) * | 2020-03-02 | 2021-03-26 | 北京航空航天大学 | Knowledge graph embedding and reasoning method introducing entity type automatic representation |
| CN111523007B (en) * | 2020-04-27 | 2023-12-26 | 北京百度网讯科技有限公司 | Method, device, equipment and storage medium for determining user interest information |
| CN112732923B (en) * | 2020-11-13 | 2022-07-12 | 哈尔滨工业大学 | Express delivery logistics service semantic extraction method based on knowledge graph |
| CN112420042A (en) * | 2020-11-19 | 2021-02-26 | 国网北京市电力公司 | Control method and device of power system |
| CN112818222B (en) * | 2021-01-26 | 2024-02-23 | 吾征智能技术(北京)有限公司 | Personalized diet recommendation method and system based on knowledge graph |
| CN114817737B (en) * | 2022-05-13 | 2024-01-02 | 北京世纪超星信息技术发展有限责任公司 | Cultural relic hot spot pushing method and system based on knowledge graph |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105868422A (en) * | 2016-06-21 | 2016-08-17 | 东北大学 | Collaborative filtering recommendation method based on elastic dimensional feature vector optimized extraction |
| CN106815307A (en) * | 2016-12-16 | 2017-06-09 | 中国科学院自动化研究所 | Public Culture knowledge mapping platform and its use method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8873813B2 (en) * | 2012-09-17 | 2014-10-28 | Z Advanced Computing, Inc. | Application of Z-webs and Z-factors to analytics, search engine, learning, recognition, natural language, and other utilities |
| US10762538B2 (en) * | 2014-04-24 | 2020-09-01 | DataSpark, PTE. LTD. | Knowledge model for personalization and location services |
| CN107122399A (en) * | 2017-03-16 | 2017-09-01 | 中国科学院自动化研究所 | Combined recommendation system based on Public Culture knowledge mapping platform |
| CN107256238B (en) * | 2017-05-23 | 2019-12-17 | 深思考人工智能机器人科技（北京）有限公司 | personalized information recommendation method and information recommendation system under multiple constraint conditions |
-
2018
- 2018-12-26 CN CN201811604829.7A patent/CN109710935B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105868422A (en) * | 2016-06-21 | 2016-08-17 | 东北大学 | Collaborative filtering recommendation method based on elastic dimensional feature vector optimized extraction |
| CN106815307A (en) * | 2016-12-16 | 2017-06-09 | 中国科学院自动化研究所 | Public Culture knowledge mapping platform and its use method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109710935A (en) | 2019-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109710935B (en) | Museum navigation and knowledge recommendation method based on cultural relic knowledge graph | |
| CN115238101B (en) | Multi-engine intelligent question-answering system oriented to multi-type knowledge base | |
| CN106649786B (en) | Answer retrieval method and device based on deep question answering | |
| CN101639857B (en) | Method, device and system for establishing knowledge questioning and answering sharing platform | |
| US8001152B1 (en) | Method and system for semantic affinity search | |
| CN111475623A (en) | Case information semantic retrieval method and device based on knowledge graph | |
| CN102955848B (en) | A kind of three-dimensional model searching system based on semanteme and method | |
| CN102663129A (en) | Medical field deep question and answer method and medical retrieval system | |
| CN106815252A (en) | A kind of searching method and equipment | |
| CN106951494A (en) | A kind of information recommendation method and device | |
| CN107402912B (en) | Method and device for analyzing semantics | |
| CN107436916B (en) | Intelligent answer prompting method and device | |
| Demir et al. | Interactive sight into information graphics | |
| WO2021179455A1 (en) | Scientometric method for measuring happiness indexes of tourist attractions | |
| CN113569023A (en) | Chinese medicine question-answering system and method based on knowledge graph | |
| CN109522396B (en) | Knowledge processing method and system for national defense science and technology field | |
| CN103927339B (en) | Knowledge Reorganizing system and method for knowledge realignment | |
| CN112115252A (en) | Intelligent auxiliary writing processing method and device, electronic equipment and storage medium | |
| CN113190593A (en) | Search recommendation method based on digital human knowledge graph | |
| JP6932162B2 (en) | Area-based item recommendation terminal device and item recommendation information provision method. | |
| US20220083879A1 (en) | Inferring a comparative advantage of multi-knowledge representations | |
| CN110110143B (en) | Video classification method and device | |
| CN114253990A (en) | Database query method and device, computer equipment and storage medium | |
| CN117420998A (en) | Client UI interaction component generation method, device, terminal and medium | |
| CN112559711A (en) | Synonymous text prompting method and device and electronic equipment |
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 |