CN105786897B - For providing the context aware body constructing method for paying close attention to information of the user based on context aware - Google Patents

Abstract

The present invention relates to the methods of the concern information based on context aware (context awareness) technological prediction user, according to the present invention, it provides a kind of for providing the context aware ontology implementation method for paying close attention to information service of the user based on context aware, wherein, it solves to provide a large amount of unnecessary search result for the retrieval request of user, to force what user filtered information needed again to be present in the problems in existing search method, in order to which based on the context aware technology Accurate Prediction information that user needs in current context and the information is supplied to user, component is for Accurate Prediction and provides the situation ontologies (context ontology) of user's information needed and travels ontology (our ontology) and define the relationship between core classes, to By semantic data mining algorithm (Semant ic Data Mining Algorithm) Accurate Prediction and user's information needed can be provided.

Description

For providing the context aware ontological construction for paying close attention to information of the user based on context aware Method

Technical field

The present invention relates to the methods of the interest based on context aware (context awareness) technological prediction user, especially It is related to being intended to provide the information that user needs in current context based on context aware technology Accurate Prediction and proposes the information The user based on context aware for supplying user pays close attention to the context aware ontology implementation method based on context aware of information service.

Moreover, it relates to which the user based on context aware pays close attention to information service providing method, wherein in order to accurate pre- User's information needed is surveyed and provided, situation ontologies (context ontology) and tourism ontology based on context aware are constructed (tour ontology) and the relationship between core classes is defined, so as to pass through semantic data mining algorithm (Semantic Data Mining Algorithm) Accurate Prediction and provide user's information needed.

Background technique

Recently, with the arrival of information age, there are various information on the network including internet, therefore, are Which type of information of accurate judgement is useful information and only selects required information and provide in bulk information, to information mistake The demand of filter technology is continuously increased.

In addition, recently, in addition to desktop computer, with the development of the technologies such as WiFi, 3G, 4G, GPS, sensor, such as pen Remember this computer (Notebook), tablet computer (Tablet PC), PDA (Personal Digital Assistant), PCS (Personal Communications Service) and smart phone (Smart Phone) etc. are in view of ambulant pervasive The development of computing technique and mobile calculation technique is quite fast.

At the same time, with the trend, the information retrieval mode under the mobile environment also occurs very big Variation, in other words, using the existing information retrieval of desktop computer and information transmitting methods be primarily adapted for use in office or Information retrieval in the fixed spaces such as family, but now with the development for the technology for manufacturing various mobile devices, information transmitting side Formula is changed into the method for considering the information retrieval of mobility and portability or can transmitting information.

Specifically, for example, pervasive web application environment is determined according to [Kim 2008] (referring to bibliography 1) Justice is what the various equipment utilization web applications such as a large number of users and desktop computer, office equipment, household appliances interacted Environment, and in the environment, the portable devices such as mobile phone can be used to obtain desired service for most users.

In addition, being compared with the information retrieval side of wired mode according to [Michalis 2012] (referring to bibliography 2) Method and using wireless mode information retrieval method the result shows that, the information retrieval method of existing fixation exists in addition to user On the whole it is in reduced trend except the period of office work or time After Hours, and utilizes the information of wireless mode Search method is in addition to the time that user sleeps, in the trend continued to increase.

As described above, the development of the information retrieval method with wireless mobile apparatus and using wireless mobile apparatus, user It can effectively retrieve information needed whenever and wherever possible, but when user search is to information needed, compared with the accurate information that user needs, More unwanted information are provided, and this is problem universal present in the existing method using desktop computer.

Especially, as described above, existing search method not only provides required concern information, therefore, it is necessary to by with Search result is filtered at family one by one, relatively complicated, moreover, the position in order to know affiliated information, also needs to retrieve map etc. again, until Until obtaining information needed, repeated retrieval need to be carried out by different classifications to identical content.

Therefore, to solve the problems, such as the search method in the existing mobile environment, need to develop situation sheet Body Model, the situation ontologies model require to provide best and personalized information inspection to be directed to the retrieval of user whenever and wherever possible Rope as a result, be based on context aware technology, predict the attention rate of user, more accurately predict user where on what is needed The search result of sample, the context aware technology work as cause using the perception users' such as user current position or periphery situation Border, and the information for being suitable for it is provided.

In addition, for the limitation in view of the mobility of the mobile terminal users such as smart phone, portability and display aspect Deng more acurrate brief search result being provided, to be suitable for the limited picture of mobile phone, to need to develop and use The semantic data mining algorithm (Semant ic Data Mining Algorithm) of context data, it is described to utilize context data Semantic data mining algorithm when user input needed for retrieval request when, situation ontologies (context according to the pre-stored data Ontology) database lookup query pattern similar or consistent with the query pattern of user, and examined using user query mode Cable system provides a user more acurrate quick search result, described to utilize user query mode search system and user search After the query pattern of request is compared, same or similar result is provided.

It is particularly preferred to ground, to overcome the retrieval request for being directed to user to show showing for a large amount of unnecessary search result The shortcomings that some retrieval modes, and provide for the retrieval request of user more acurrate brief as a result, it is desirable to provide new information Search method, the new information retrieval method situation perceived according to the search result of user and information that customization is provided it is same When, using the GPS and map function for being built in the mobile devices such as smart phone, provided together with search result to the current of user The information of position and the position of user's information needed to make user that can rapidly and accurately find information needed, but does not propose also It can satisfy the device or method of the requirement.

[bibliography]

1、Kim H.W.,Nam J.Y.,Lee H.,Shin D.M.,"A Device Coordinat ion Mechanism with Context Ontology for Ubiquitous Web Appl icat ion Environment",Journal of Engineering&Technology,Vol.17,2008.

2、Michal is K.,"An Introduction to Android",Tutorial:hy439&hy539, 2012.

3、Uschold M.F.,Jasper R.J.,"A Framework for Understanding and Classifying Ontology Appl icat ions",Proceedings of the IJCAI-99Workshop on Ontologies and Problem Solving Methods(KRR5),pp.1-12,1999.

4、Fischer G.,Ostwald J.L.,"Knowledge Management:Problems,Promises, Real i t ies,and Chal lenges",IEEE Intel l igent Systems,Vol.16,No.1,pp.60- 72,2001.

5、Hwang H.S.,Lee J.Y.,"A Study of a Knowledge Inference Algor i thm us ing an Associat ion Mining Method based on Ontologies",Journal of Korea Mul t imedia Society,Vol.11,No.11pp.1566-1574,2008.

6、Mostefaoui S.K.,Hirsbrunner B.,"Towards a Context-Based Service Compos i t ion Framework",Proceedings of the Internat ional Conference on Web Services,pp.42-45,ICWS'03,Las Vegas,Nevada,23-26June 2003.

7、Yi lmaz N.,Alptekin G.,"An Ontology-Based Data Mining Approach for Strategic Market ing Decis ions",The 10th Internat ional Conference on Cybernet ics and Informat ion Technologies,Systems and Appl icat ions:CITSA 2013.

8、Kim J.H.,Kim H.C.,"Informat ion Retr ieval System for Mobi le Devices",Journal of the Korean Society of Mar ine Engineer ing,Vol.33,No.4, pp.569-577,2009.

9、Park H.S.,Park M.H.,Cho S.B.,"Information Recommendat ion in Mobile Environment using a Multi-Criteria Decision Making",Journal of KI ISE.Comput ing Pract ices and Letters,Vol.14,No.3,pp.306-310,2008.

10、Seo H.S.,Lee S.Y.,"User Behavior Pattern Inference Model for Context-Awareness based Personal ization Recommended Services",Thes is (Master's degree),Kongju Nat ional Univ.Graduate School:Computer Science Department,2012.

11、Jeong J.E.,"Ontology based Preprocess ing Scheme for Mining Data Streams from Sensor Networks",Intelligence Information Research,Vol.15,No.3, pp.67-80,2009.

12、Ko S.K.,Choi S.M.,Eom H.S.,Cha J.W.,Cho H.C.,Kim L.H,Han Y.S.,"A Smart Movie Recommendation System",Human Interface and the Management of Information.Interacting with Information Lecture Notes in Computer Science, Vol.6771,pp.558-566,2011.

13、Sarkaleh M.K.,Mahdavi M.,Baniardalan M.,"Des igning a Tour ism Recommender System Based on Location,Mobile Device and User Features in Museum",International Journal of Managing Information Technology(IJMIT), Vol.4,No.2,May 2012.

14、Abraham T.,"Knowledge Discovery in Spatio-Temporal Databases", School of Computer and Information Science,University of South of Australia, Ph.D dissertation,1999.

15、Gu M.S.,Hwang J.H.,Ryu K.H.,"Weigted-FP-Tree Based XML Query Pattern Mining",6th International Conference,ADMA 2010,Vol.6440,pp.417-428, 2010.11.

16、Chen,L.Bhowmick,S.S.Chia,L.T.,"Mining Maximal Frequently Changing Subtree Patterns from XML Documents",In Proceedings of the 6th International Conference on Data Warehousing and Knowledge Discovery(DaWaK),Zaragoza,Spain, 2004.

17、Chen,L.Bhowmick,S.S.Chia,L.T.,"FRACTURE-Mining:Mining Frequently and Concurrently Mutating Structures from Historical XML Documents",Elsevier Science Journal:Data&Knowledge Engineering Volume:59Issue:2pp.320-347,2006.

18、Han,J.Pei,J.Yin,Y.,"Mining Frequent Patterns without Candidate Generat ion",Proceedings of the 2000ACM SIGMOD international conference on Management of data,2000.

19、Hunyadi D.,"Performance comparison of apriori and FP-growth algorithms in generating association rules",Proceedings of the 5th European conference on European comput ing conference pp.376-381,2011.

20、http://www.android.com/

21、http://www.jetbrains.com/webstorm/

Advanced technical literature

1, Korean Patent Laid the 10-2013-0097476th (2013.09.03.)

2, Korean registered patent gazette the 10-1009638th (2011.01.13.)

3, Korean registered patent gazette the 10-0868331st (2008.11.05.)

4, Korean registered patent gazette the 10-0688087th (2007.02.22.)

Summary of the invention

The present invention is proposed to solve existing issue as described above, and thus the purpose of the present invention is to provide one kind For providing the context aware ontology implementation method for paying close attention to information service of the user based on context aware, wherein existing in order to solve There is the problem of retrieval mode of technology, based on the attention rate of context aware technological prediction user, constructs more acurrate prediction user and exist Where on which type of is needed retrieve situation ontologies (context ontology) model, and by according to the situation Ontology model predicts the concern information of user, only provides information relevant to user's concern for the retrieval request of user, thus It can provide more acurrate brief search result, wherein the retrieval mode of the prior art has the disadvantage that, for user Retrieval request a large amount of unnecessary search result is provided, so that user be forced to filter information needed again.

In addition, paying close attention to letter for providing the user based on context aware it is a further object of the invention to provide a kind of Cease service context aware ontology implementation method, wherein in order to overcome the mobile phones such as smart phone limited picture and The defect of few memory provides more quasi- using semantic data mining algorithm (Semantic Data Mining Algorithm) Really brief search result, the semantic data mining algorithm utilize situation ontologies.

Another purpose again of the invention is, provides a kind of user's concern information clothes for providing based on context aware The context aware ontology implementation method of business, wherein can not only be easy to find information needed to only pass through primary retrieval Content, but also position can be found, while providing the search result for the retrieval request of user, using being built in The GPS and map function of mobile device, provided together with search result user current location and user needed for search result Position.

To reach the purpose, according to the present invention, a kind of context aware ontology implementation method, the context aware sheet are provided Body (Context Awareness Ontology) implementation method realizes retrieval service, wherein in order to solve the inspection of the prior art The problem of cable system, based on context aware (Context Awareness) technology only using the concern information of the user as inspection Hitch fruit provides, to can provide more acurrate brief search result, the inspection of the prior art compared with existing searching system The shortcomings that cable system be the retrieval request for user because unnecessary information is also provided as search result to user, institute So that user filters information needed again and selects, the context aware ontology implementation method includes: that user pays close attention to Information Number According to library construction step, the information of the region-of-interest for the user is collected, to construct database；Ontology inference step, It is configured the relationship and attribute that are stored in the user and pay close attention between each data of information database, in the user The user, which is constructed, in concern information database construction step pays close attention to information database；And ontology database construction step, Ontology database is constructed according to the relationship for each data being arranged in the ontology inference step and attribute.

Here, the contextual information of the user is expressed as including user identifier U (User by the ontology Identification), spatial information L (Location), temporal information T (Time) and information on services relevant to the user The space contextual information O={ U, L, T, S } when having four-dimensional of S (Service).

In addition, the temporal information is the information divided by pre-set certain time spacing, the spatial information is The general location information of the coordinate (xi, yi) of belonging euclidean (euclidean) spatially, and the coordinate is distinguished Indicate the longitude and latitude value in the area that the user is presently in.

At the same time, in the database sharing step, as the concern information of the user, by each area collection pair Each area is appointed as subclass (subclass) as most higher level's class designated root class (root class) by the information on sightseeing ground, By to described each region sightseeing ground title be appointed as the subclass to each area, by each sightseeing ground facility or Subclass to the ground title of going sightseeing is appointed as to the affiliated detailed description for going sightseeing ground, each department or the sightseeing are set Synonymous part of speech (equivalent class) relationship retrieved for synonym or setting consonant is set, to construct composition classification knot The sightseeing database of structure.

Here, the sightseeing database includes multiple domains (field), the multiple domain (field) include id, add_1, Add_2, add_3, add_4, add_5, tour_s ite, syn, lat, long, it is the domain add_1, the domain add_2, described The address in the domain add_3, the domain add_4 and each ground of going sightseeing of the adde_5 domain representation, the domain tour_site is stored with sight Light ground title, the syn are the domain for handling synonym or consonant word, and the domain lat and the domain long respectively indicate and set Set latitude and longitude required when map.

Especially, the ontology inference step according to by be gone sightseeing area title constitute " Place ", by each " Attraction " that sight-seeing resort title in area is constituted, by each sightseeing ground building or remains, traces, seashore And restaurant " Resourc " and " Activity " that constitutes determines the attributes of each data, " Activity " includes table Show that " the Cultural activity " of cultural activity, " the Scenery view " of the behavior for indicating ornamental landscape, expression go out to eat " the Dining out " of the behavior of meal, to indicate the practical action that the user is implemented on each ground of going sightseeing.

Here, the relation on attributes of " Place " and " Attraction " by described in is specified in the ontology inference step For " hasAttraction " and " IsAttractionOf ", the attribute of " Attraction " and " Resource " by described in Relationship is appointed as " hasResource " and " IsResourceOf ", and by described in " Activity " and " Resource " Relation on attributes is appointed as " hasActivity " and " IsActivityOf ".

In addition, there are when multiple relations on attributes in the data for constituting each class, being considered as tool in the ontology inference step There is multiple attributes relationship.

At the same time, it in the ontology inference step, is executed using Protege_4.0.2version.

Especially, following processing is executed in the ontology database construction step: utilizing OWL (network ontology language, Web Ontology Language), RDF (resource description framework, Resource Description Framework) and RDFS (money Source Description framework architecture specification, Resource Description Framework Schema Specification) building institute State ontology database.

In addition, according to the present invention, providing a kind of user's concern information service providing method based on context aware, utilizing By the context aware ontology of the documented context aware ontology implementation method building, user is only provided as search result Concern information therefore, be suitable for including intelligence to provide more acurrate brief search result compared with existing search method The mobile environment of mobile phone or tablet computer.

At the same time, according to the present invention, a kind of user's concern information service system based on context aware is provided, is utilized By the context aware ontology of the documented context aware ontology implementation method building, user is only provided as search result Concern information therefore, be suitable for including intelligence to provide more acurrate brief search result compared with existing searching system The mobile environment of mobile phone or tablet computer.

As described above, according to the present invention, providing a kind of for providing user's concern information service based on context aware Context aware ontology implementation method, wherein the attention rate based on context aware technological prediction user, and construct for more acurrate pre- User is surveyed in the situation ontologies model for where which type of needing retrieve, and by according to the situation ontologies model prediction The concern information of user only provides user's information of concern for the retrieval request of user, to can provide more acurrate brief The shortcomings that search result, therefore the problem of be able to solve the retrieval mode of the prior art, the retrieval mode of the prior art, exists In for a large amount of unnecessary search result of retrieval request offer of user, so that user filters information needed again.

In addition, according to the present invention, providing a kind of for providing the situation for paying close attention to information service of the user based on context aware Perceive ontology implementation method, wherein semantic data mining algorithm (the Semantic Data using situation ontologies can be passed through Mining Algorithm) more acurrate brief search result is provided, so as to overcome the mobile phones such as smart phone The defect of limited picture and few memory.

At the same time, according to the present invention, it provides a kind of for providing user's concern information service based on context aware Context aware ontology implementation method, wherein while providing the search result for the retrieval request information of user, using interior Be placed in the GPS and map function of mobile device, provided together with search result user current location and user needed for retrieval As a result position, to only can not only be easy to find the content of information needed by primary retrieval, additionally it is possible to search in place It sets.

Summary of drawings

Fig. 1 is the schematic diagram that the characteristics of expression through table to existing data digging method and advantage and disadvantage are compared；

Fig. 2 is to indicate the flow chart of the specific process for being appointed as geographic coverage with going sightseeing and constructing situation ontologies；

Fig. 3 is for utilizing the feelings for being located at the ground of going sightseeing in each area in order to provide the user with the service to ground of going sightseeing Border perception data constructs sightseeing ontology, with this provides the user with the service to ground of going sightseeing and carries out the process schematic of inference；

Fig. 4 is the generalized schematic of the part-structure in the hierarchical structure for indicate sightseeing ontology shown in Fig. 3；

Fig. 5 is for the mistake that the attribute value and relationship that constitute the data value of each class are arranged to carry out ontology inference The schematic diagram that journey is illustrated；

Fig. 6 is the schematic diagram for indicating to constitute the multiple attributes relationship between the data of each class；

Fig. 7 is the process generalized schematic that ontology inference is carried out using protege；

Fig. 8 is the partial schematic diagram for indicating the sightseeing database constructed according to embodiments of the present invention；

Fig. 9 is the partial schematic diagram for indicating the sightseeing database table constructed according to embodiments of the present invention；

Figure 10 is the schematic diagram for indicating sightseeing body construction according to an embodiment of the present invention；

Figure 11 is the schematic diagram for indicating to be unfolded the body construction of the add_2class of sightseeing ontology shown in Fig. 10；

Figure 12 is the schematic diagram for indicating to be unfolded the body construction of the add_3class of sightseeing ontology shown in Fig. 10；

Figure 13 is the schematic diagram for indicating to be unfolded the body construction of the add_4class of sightseeing ontology shown in Fig. 10；

Figure 14 is to be indicated to utilize OWL (Web Ontology using protege_4.0.2 according to the embodiment of the present invention Language), the schematic diagram of the part of the tour.owl ontology source code (ontology source) of RDF and RDFS building；

Figure 15 is to indicate according to an embodiment of the present invention to pay close attention to information service for providing the user based on context aware The integrally-built generalized schematic of context aware ontology implementation method；

Figure 16 is to respectively indicate XML query scheme-tree (Query pattern tree) and the XML query subtree with root The schematic diagram of (Query Rooted subtree)；

Figure 17 is the structural schematic diagram for respectively indicating query pattern tree and the subtree that takes place frequently with root；

Figure 18 is the process and each thing for indicating the prefix (prefix) of shared frequent pattern tree (fp tree) (FP-Tree) building process The generalized schematic of the process of business (Transact ion)；

Figure 19 is during executing the progress data mining of FP-growth algorithm, and building is used as any project (item) M condition (conditional) frequent pattern tree (fp tree) (FP-Tree) process schematic diagram；

Figure 20 is repeatedly the process of execution search criterion mode as a result, being arranged by condition pattern of the table to acquisition Schematic diagram；

Figure 21 is the exemplary diagram that the Korean inquiry for executing user is made as XML form；

Figure 22 is the schematic diagram for indicating the Book DTD Tree as the XML document made based on XQuery；

Figure 23 is the frequent item set (transaction that each affairs are indicated in sightseeing ontology database Frequent itemset) schematic diagram；

Figure 24 is the frequent pattern tree (fp tree) (FP- using the sightseeing ontology database with the structure that takes place frequently occurred by each department Tree) the partial schematic schematic diagram of structure；

Figure 25 is to respectively indicate to construct condition on the basis of the Ma Luolu in the whole world (Global) FP-tree and area of going sightseeing (Conditional) schematic diagram the case where FP-tree；

Figure 26 is the schematic diagram for indicating the part of original code of FP-growth algorithm；

Figure 27 is that in FP-growth algorithm to be applicable in minimum support be 0.4 and the result that carries out data mining is shown It is intended to；

Figure 28 is to be applied to the method for mode matching using semantic data mining algorithm based on mobile information inspection The overall structure generalized schematic of cable system；

Figure 29 is that a series of outline for the treatment of processes performed when receiving the retrieval request to ground of going sightseeing from user shows It is intended to；

Figure 30 is to illustrate when receiving retrieval request from user, by analyzing the requirement item of user and obtaining user's Query pattern, and by the mode with the consistent mode of the query pattern of analysis and the ontology database for being stored in searching system Structure is compared, and retrieves the schematic diagram of the process of mode configuration similar or identical；

Figure 31 is the entirety that information retrieval method is paid close attention to according to the user of the embodiment of the present invention excavated using semantic data The generalized schematic of structure；

Figure 32 is for realizing the schematic diagram of the part of original code of the initial picture of retrieval window；

Figure 33 is the event-action (event realized with script function (Script function) to index button Action source code schematic diagram)；

Figure 34 is to execute for providing the part source generation of the map button behavior (map button action) of Map Services The schematic diagram of code；

Figure 35 is the schematic diagram according to the composition of the key frame of the mobile message searching system of the embodiment of the present invention；

Figure 36 is the ground zones values respectively indicated to retrieval window input crop field wide area city and shows institute when clicking map button The composition of the map in possession area and according to retrieval window input input value show introduce relative region restaurant or go sightseeing The schematic diagram of the composition of the picture in area；

Figure 37 is to realize through latitude and the present position values of longitude automatic sensing user to show that user works as simultaneously The map and latitude in preceding locating area and the map of the map of longitude realize the schematic diagram of source code；

Figure 38 is the schematic diagram for the result screen realized by source code as shown in figure 37；

Figure 39 is that data digging will be executed according to being applicable in FP-growth algorithm in ontology database of the embodiment of the present invention The interpretation of pick process is the schematic diagram of table；

Figure 40 is the reality that will be applicable in FP-growth algorithm in existing relevant database and execute data mining process Test the schematic diagram that result arrangement is table；

Figure 41 be according to the embodiment of the present invention will be as the minimum of ontology data and existing relevant database be supported The change procedure for the memory usage amount that the variation of degree critical value is changed is expressed as the schematic diagram of chart；

Figure 42 be according to the embodiment of the present invention will be as the minimum of ontology data and existing relevant database be supported The quantity variation for the frequent mode that the variation of angle value is changed is expressed as the schematic diagram of chart；

Figure 43 be according to the embodiment of the present invention will be as the minimum of ontology data and existing relevant database be supported The variation for the execution time that the variation of degree is changed is expressed as the schematic diagram of chart；

Figure 44 be according to the embodiment of the present invention will be with the change of the affairs of ontology data and existing relevant database The variation for changing the changed execution time is expressed as the schematic diagram of chart；

Figure 45 is according to the embodiment of the present invention to the affairs with ontology database and existing relevant database The variation for changing changed memory usage amount is compared and is expressed as the schematic diagram of chart；

Figure 46 is by the variation table for the accuracy for indicating recall precision with the variation of the affairs as user query mode It is shown as the schematic diagram of chart；

Figure 47 is in existing portal when the picture of the search result shown when retrieving window and inputting " chicken Longshan " The schematic diagram in face；

Figure 48 is to indicate in mobile message searching system according to an embodiment of the present invention, in retrieval window input " chicken dragon The schematic diagram of the picture of shown search result when mountain "；

Figure 49 is to indicate in mobile message searching system according to an embodiment of the present invention, is shown when inputting search key The schematic diagram of the picture of the map of detailed description relevant to search result and surrounding area.

Specific embodiment

In the following, in conjunction with attached drawing to of the invention for providing the situation sense for paying close attention to information service of the user based on context aware Know that the specific embodiment of ontology implementation method is illustrated.

Here, content described below be a kind of embodiment for carrying out the present invention, Ying Liuyi the present invention is not limited to Embodiment content as described below.

In addition, illustrating below in the explanation of the embodiment of the present invention to simplify, or class identical as the content of the prior art Seemingly, or skilled addressee readily understands that and the part implemented, the detailed description that carries out to it will be omitted.

Words sentence is talked about, and as described below, pays close attention to information clothes for providing the user based on context aware the present invention relates to a kind of The context aware ontology implementation method of business, wherein in order to solve the problems, such as the retrieval mode of the prior art, be based on context aware skill Art predicts the attention rate of user, and building is for more acurrate prediction user in the situation ontologies for where which type of needing retrieve (context ontology) model, and by the concern information according to the situation ontologies model prediction user, for user Retrieval request information relevant to user's concern is only provided, it is described existing to can provide more acurrate brief search result The retrieval mode of technology has the disadvantage that the retrieval request for user provides a large amount of unnecessary search result, thus So that user filters information needed again.

At the same time, the present invention relates to a kind of for providing the situation sense for paying close attention to information service of the user based on context aware Know ontology implementation method, wherein in order to overcome the limited picture of the mobile phones such as smart phone and the defect of few memory, More acurrate brief retrieval knot is provided using semantic data mining algorithm (Semantic Data Mining Algorithm) Fruit, the semantic data mining algorithm utilize situation ontologies.

Especially, the present invention relates to a kind of for providing the context aware for paying close attention to information service of the user based on context aware Ontology implementation method, wherein in order to only pass through the content that primary retrieval can not only be easy to find information needed, additionally it is possible to look into Position is found, while the search result for the retrieval request information of user is provided, utilizes the GPS for being built in mobile device And map function, the current location of user and the position of search result needed for user are provided together with search result.

Here, being used to provide the context aware for paying close attention to information service of the user based on context aware to according to the present invention Before the specific embodiment of ontology implementation method is illustrated, to being suitable for the invention ontology (Ontology), context aware The basic conception of (Context Awareness) and data mining algorithm (Data Mining Algorithm) is illustrated.

Firstly, ontology (Ontology) refer to in particular range information carry out generalities, with logical construction classification or Defined notion, thus the method indicated in the form of hierarchical structure, in the prior art, according to [Uschold 1999], [Fischer 2001] and [Hwang 2008] etc., ontology are the data for analytical integration by dispersions such as network and suppliers Data presentation technique, relative to shared generalities be defined as standardization and specific detail, it is however generally that, for structure Ontology is built to propose to as RDFS (resource description framework framework specification, the Resource for indicating the element of ontology Description Framework Schema Specification), RDF (resource description framework, Resource Description Framework), (reference of the technology contents of OWL (network ontology language, Ontology Web Language) Bibliography 3 is to bibliography 5).

In other words, RDFS is as the most basic language for indicating ontology, and definition is for using the basic of RDF applicational language Vocabulary, and useful language of the RDF as the information and metadata that can indicate all kinds in data model, by resource and Sentence is constituted.

Specifically, RDF sentence is by subject (subject), narration (predicate (property)), purpose (object (value)) it constitutes, and OWL (network ontology language, Web Ontology Language) is for first as RDFS Define ontology in the semantic network made based on the Ontology Language of phase and according to the standard body language that W3C is proposed, to indicate The hierarchical structure of ontology can indicate the hierarchical structure up and down between object class and classification, any attribute can be defined by the user simultaneously The attribute of definition is distributed into classification, therefore, compared with RDFS more standard, and logic class can be carried out in relationship between objects It pushes away, equally, with RDF sentence expression information.

Here, the embodiment of the present invention that will illustrate below has used institute also for building sightseeing ontology database State RDFS, RDF, OWL etc..

Then, as follows to the conceptual illustration of context aware (Context Awareness): contextual information can be such as user The same personal information of current active, can be the technical information as the equipment in currently used, be also possible to as Temperature, position or time the same environmental information.

In other words, it is to be perceived by the collection and exchange of the contextual information, and pass through that context aware is individualized It explains and the treatment processes such as inference provides a user the customization service for being suitable for situation, for this purpose, the various contextual informations on periphery need Obtain modeling, collection and distribution appropriate.

Therefore, in order to provide context-aware services, the contextual information being collected into need to be analyzed and be pushed away in conjunction with user information By the high-level information of the one aspect at expressible life, as described above, being stored with information appropriate and system effectivelys share Method be determine context aware system success or not important technology.

At the same time, recently, it with the increase for the user for using mobile device, answers fast-changing environment is suitable for It is continuously increased with program and the demand of service, but is difficult to construct the context aware system for meeting the requirement so far to overcome The problem of system, as semantic situation performance, situation inference, knowledge sharing, context classification method, it is proposed that utilizing The environmental model (context model) based on ontology of OWL.

Specifically, in the prior art, to provide the context-aware services that can carry out various situation inferences, proposing SOCAM (service-oriented occasion incudes middleware, Service-Oriented Context-Aware Middleware), is changed Sentence is talked about, such as according to [Mostefaoui 2003], it is proposed that in conjunction with Context-aware computing (context-aware comput Ing) and agency (agent) technology CB-Sec (Services Composition based on situation, Context-Based Service Composition) frame (framework) (referring to bibliography 6).

It comprises history (History) as a result, to overcome situation in most of context aware application program The shortcomings that parameter (context parameter) only assert position (location) and ignores the aspect of time (temporal), It include following 5 kinds of elements: user context (user context) in existing context aware comprising the effect of user, Identity, age, position, content of concern etc.；It calculates situation (Comput ing context) comprising network connection and periphery Printer or display, work station etc.；Time situation (Time context) comprising the times such as day, week, the moon, season；Object It manages situation (Physical context) comprising weather, temperature etc.；And situation history (context history), it will User context calculates situation and physics situation record certain time, so as to be effectively applied in the service of composition.

Then, data digging method is described as follows: in order to be applicable in semantic data method for digging, is also can be used existing Text mining method, but in order to utilize the referred to as sightseeing ontology database for being suitable for the embodiment of the present invention being described below Ontology executes data mining, here, saying to the data digging method of the XML document about the basic structure for constituting ontology It is bright.

In other words, referring to Fig.1, Fig. 1 be indicate by table to existing data digging method the characteristics of and advantage and disadvantage into The schematic diagram that row compares.

As shown in Figure 1, in the prior art, having studied to the data digging method of various XML datas and to XML query mould The various methods that formula is excavated, but there are advantage and disadvantage for each method.

In addition, in present searching system, providing excessive inspection because huge data exist according to [Hwang 2008] Hitch fruit or the case where search result cannot be provided, therefore, for the defect for making up the searching system, it is proposed that will pass through The relevant knowledge of OWL ontology restriction condition and related method for digging institute inference shows as SWRL inference language, to pass through jess Engine finds new knowledge and supports the algorithm effectively retrieved, and in other words, building is paid close attention to according to the individual of domain (domain) Ontology provides related and inference information mistake when illustrating based on related and ontology retrieval information by taking personal focused data as an example Journey (refers to bibliography 5).

At the same time, according to [Yilmaz 2013], the data mining side based on ontology for utilizing Apriori algorithm is proposed In other words method for the operating system for determining consumers, proposes the frame based on data digging method, in building ontology And questionnaire survey is carried out after measuring preference to user, data are transmitted, analysis, which clusters, simultaneously to be divided using dependency rule algorithm It analyses as a result, at this point, being to construct ontology and use protege, and Apriori algorithm executes following two steps: searching the item that takes place frequently Mesh set, and the project set is divided into two kinds of precedent (antecedent) and result (consequent), so that searching can Meet the rule of confidence level (confidence) (with reference to bibliography 7).

Then, the information retrieval system of mobile environment is described as follows: recently, with the development of prevailing technology, as right The research for providing a user the context aware technology of useful service to the situation for perceiving user and provides the feelings of useful information Border perception recommendation service is studied, but the recommendation information based on situation is only recommended as one by existing context aware recommendation service The information for the concept level passed through, therefore, when the information recommended is not suitable for the current context of user, user need to be in recommendation Information needed is directly searched in information, to need to expend many effort of user and time.

At the same time, recently, with the development of mobile internet environment, become living using the Internal retrieval of terminating machine Jump, and in order to effectively input search key, proposing the method in terms of considering the hardware capabilities such as keyboard and consider to utilize user Collection, storage keyword etc. software function in terms of method.

Here, ontology model can be used as storing and managing the mobile devices such as GPS, illuminance transducer, user information Various contextual informations purposes, and mobile device is obtained according to the type for the hardware sensor for being included in equipment to the position of user It sets, various contextual informations of environment etc., and is assigned as ontology model, understand the meaning between contextual information, from And it is processed into the information of required form or is used for situation inference.

In addition, mobile environment have the feature different from general cable environment, in other words, first, mobile environment not It is limited by time and space, therefore, user carries mobile device and retrieves information needed whenever and wherever possible；The second, mobile environment is non- Often individualized, most mobile device is person ownership, not according to information needed for personal interest and region-of-interest Together, therefore, according to personal feature, even if identical content on surface, the content actually inquired is possible to different；Third, Because mobile device be all it is small-sized, input, there are many limitations for output facet, therefore, in mobile environment, compared with Many pages shows by small picture, information needed for only extracting user show more effectively；4th, general and Speech, the data transmission rate of mobile environment is much lower compared with cable environment, but price is higher；5th, in general, mobile device is deposited In the limitation of memory and processing capacity, therefore, it is impossible to handle a large amount of information in mobile device itself.

Therefore, it is necessary to consider the mobile environment characteristic and provide the information retrieval system for being suitable for mobile environment.

In other words, it according to [Kim 2009], the characteristics of considering the mobile environment, proposes document summarization and user Information (user profile) is suitable for the mobile message retrieval model of member retrieval (meta search) model, and the document is general The technology of including can be relieved the problem of low transfer rate in small picture, and user information can be in mobile environment using personalized Information retrieval model (refers to bibliography 8).

At the same time, different according to situation for the preference of information recommendation service, and in order to provide information recommendation Service, first it is to be understood that the contextual information of user.

In other words, according to [Park 2008], it is proposed that consider the recommendation system of the preference of most users in mobile environment The recommendation in restaurant is united and is suitable for, and in order to be monitored in mobile environment to the preference of individual user, use pattra leaves This network, and restaurant is recommended in order to which the preference acquisition of the most users in most cases based on non-individual user is multiple partially Good degree, the AHP (referring to bibliography 9) used as multi-standard decision-making technique.

Recently, as most of user possesses mobile device, mobile device not only acts as the channel of accessible personal information Effect, moreover, the work of the tool of the environmental information of collectable user can also be played because user takes take with oneself always With, and the information being collected into this way can be efficiently used in the information recommendation for user, therefore, the inference of mobile environment be also with The relevant important field of research of information recommendation.

At the same time, in order in environment sensing environment provide personalized recommendation service, need to quickly analyze collected by The purpose of environmental information and valid inference user, but there are numbers according to different situation environment for the information being collected by mobile device According to difference be therefore not suitable for directly being applicable in existing inference algorithm, and need to be suitable for the effective algorithm of mobile environment.

In other words, according to [Seo 2012], caused by reducing omission or mistake because of information etc. to the maximum extent Loss, classifies to action model using Naive Bayes Classifier, and in order to effectively learn the tendency of user and inference row It moves purpose and has used method for mode matching (with reference to bibliography 10), in addition, being described according to [Jeong 2009] with each The exploitation of kind of sensor and the building of sensor network, for collect the environmental data of particular space and find more useful information and The research (referring to bibliography 11) of knowledge and utilization data mining (Data mining) method.

At the same time, according to [Ko 2011], the recommendation system based on type of relationship (Genre correlation) is proposed System is believed to collect preference from a large number of users using collaborative filtering method (Collaborative filtering method) Breath, so that (bibliography 12 may be referred to) for the automatic Prediction of the concern item of user, and [Sarkaleh2012] In propose following recommender system: because such as laptop, mobile phone, PCS (PCS Personal Communications System, personal Communication system), the movement as PDA (palm PC, personal digital assistant) etc. sets It is standby to have the characteristics that mobility, independence, individual, accessibility, therefore made using the equipment in correct time and field Study it is individualized, and but also learning Content is rich and varied, and utilize GPS (system of geographic location, Geographic Position System) function can obtain the information (with reference to bibliography 13) in various regions needed for user.

As described above, in the prior art, proposing the concern with the information for only providing necessity and concern to user The relevant various methods of information recommendation system, but in nearest mobile environment, existing information retrieval method exists as follows Problem.

Specifically, more existing data retrieval mode and the data retrieval mode of current moving area ought be commonly used to When, it is possible to find maximum difference is mobility, and in other words, existing data retrieval method is because most of use desktop Brain is applicable in so being fixed on one place and realizing all retrievals and data, but in the mobility and work for considering today's society When dynamic property, the efficiency of the existing data retrieval mode is at a fairly low, and can be used as and lag behind the most advanced of today's society The mode of technology.

Therefore, in order to overcome existing information retrieval mode the shortcomings that, the information retrieval to mobile environment is suitable for The demand of method is continuously increased, in addition, can not construct can examine because existing information retrieval mode does not consider mobility Consider the system of the current location of user, but with the development of recent mobile technology, can will search the current of user using GPS function The context aware method of position is incorporated in information retrieval system, therefore, is provided following service in the present invention: being passed through the skill Art executes the current location of inquiry using user to transmit accurate information.

In other words, for example, utilizing situation to provide a user more accurate information transmitting service based on context aware Perception ontology collects the sightseeing data information in each area, to construct sightseeing ontology, and can provide and is suitable for user and works as cause The information service in border.

At the same time, when inputting keyword in existing information retrieval system and obtaining search result, compared with user institute There are more unrelated items in the accurate information wanted, thus the problem of unnecessary huge search result is provided, therefore, In the present invention, context aware body of data is constructed, and in order to solve described problem in order to which the inquiry offer to user is more quasi- Really quickly service, has used semantic data method for digging.

In other words, for example, constructing sightseeing ontology and storing the query pattern that user often inquires, when user inquires When, using the matching process with stored query pattern and user query, provide a user consistent or similar with user query Query pattern, thus only provide user's information actually required.

Especially, existing information retrieval system only provides text information, and therefore, the present invention provides Map Services (Map Service), taken using the map such as the Google Maps (Google map) that the mobile device including smart phone is supported Business, the particular geographic location information of information needed for only can quickly and easily finding even user by primary retrieval.

Then, to solve the problem of the prior art, the detailed process of the ontology of context aware information is utilized to building It is described in detail.

Firstly, being described as follows to the basic conception of context aware ontology (Context Awareness Ontology): when Space Move Mode is to be realized to individualize according to the position characteristic of customer by the mode position of mobile object, and make suitably Interior perhaps service Spatial Rules when becoming possible are provided, and it is described when spatial relationship to find to when spatial object it is relevant Between event causality (causal relationship) tool has very important significance, therefore, can by simultaneously consider when Between and the ontology in space can provide service appropriate to the action of user.

As a result, in the present invention, it designs and is conducive to structuring contextual information and can indicate correlation and part feelings The ontology of the information in border, and according to the ontology of the embodiment of the present invention be with as time and spatial information and service etc. The relevant information of the contextual information of user is expressed as O={ U, L, T, S }.

Here, U (User identification) is user identifier, L (Location) is spatial information, T (Time) For temporal information, and S (Service) indicates service regulation related to user.

In other words, the when space contextual information of user U include the identification information of user, temporal information, spatial information and The four-dimensional informations such as information on services.

Specifically, the when spatial information of the room and time of user is the contextual information for indicating user for identification, and For distinguishing the benchmark of the service log according to different situations, and the location information of user by actual event occur it is effective when Between coordinate spatially generalization value constitute.

At this point, temporal information is the information divided by certain time spacing, spatial information is belonging Euclidean space On coordinate (xi, yi) general location information.

In addition, the generalization of position carries out generalization by certain area (zone) to the coordinate value in space, to identify use Family and the position of object, here, space coordinate respectively indicates the longitude and latitude value in the area being presently in.

For example, the space coordinate of user is general if user is currently located at crop field wide area Shizhong District ginkgo hole in the street The zone locations value of the position in belonging crop field wide area Shizhong District ginkgo hole is turned to, so as to utilize the latitude of the current location of user Angle value and longitude.

Then, the process of ontology inference process needed for building ontology database and building sightseeing ontology is illustrated.

Firstly, Fig. 2 is to indicate the specific process for being appointed as geographic coverage with going sightseeing and construct situation ontologies referring to Fig. 2 Flow chart.

As shown in Fig. 2, the process of building situation ontologies is, firstly, determining the ontology to be constructed for building ontology Domain (S21 step).

In other words, for example, in the case where providing the information to ground of going sightseeing, sightseeing ground information needed for searching user, To collect the information by ground of going sightseeing belonging to each area.

Then, the class (S22 step) to each main concept is constituted, at this point, each class is by superclass (superclass) It is constituted with the hierarchy of the compositions such as subclass (subclass).

In other words, for example, as the specified root class (root class) for being known as tour_sites of most higher level's class, under it Each area is appointed as subclass (subclass) and generates lower layer's class by face, and more accurate subclass is arranged to each area and closes System, to construct the sightseeing ontology for constituting hierarchical structure.

Then, have the characteristics which kind of relationship and which kind of using each data of attribute definition for the data value for constituting each class (S23 step), in other words, for example, by the characteristic in the area for constituting each class or the famous-eat being included in affiliated area or saying The bright content for going sightseeing ground is appointed as the relation on attributes of each class.

Then, to the sightseeing of belonging each class title utilizes characteristic specified attribute value possessed by the sightseeing ground, from And it generates and is known as ground example (instances) (S24 step) of going sightseeing.

Then, consistency (cons istency) is executed to check, check the sightseeing ontology that newly constructs whether with it is existing Ontology is upper consistent (S25 step) in axiom or attribute etc., and the ontology constructed in this way is used for the information retrieval of user.

Therefore, by executing the process, when user's sightseeing needed for the retrieval of current location when, in order to recommend with Sightseeing ground appropriate, classifies by different regions with going sightseeing, to construct sightseeing ontology on position required by family.

Then, illustrate for the building sightseeing ontology and the process of all kinds of attributes and relationship of inference sightseeing database.

In general, user wishes not influenced by the location of oneself or surrounding enviroment, it is in any position or any Best service is obtained in environment, at this point, service is to the available function as equipment, position, execution condition etc. Information describes after being virtualized and realizes the unit of hierarchical structuring.

In addition, it is necessary to if be claimed new in stored service regulation set there is no the service that composite users require Service or combine existing service to be provided, and in order to provide a user such service, need to intrinsic each place and Positioned at each place sightseeing and each sightseeing ground in the service available content such as restaurant, shopping center, seashore, boulevard into Row inference, to specify intrinsic all kinds of specified attributes.

In addition, if being needed in existing stored service regulation set there is no the service for being suitable for user's requirement It generates new service or the existing service of combination and provides, in order to provide a user the service, to can provide in the body Each place and sightseeing positioned at each place, and restaurant shopping center living, seashore, mountain-climbing etc. positioned at each ground of going sightseeing The content of service need to specify the attribute of intrinsic each class by inference.

In order to illustrate the inference process of the sightseeing ontology, firstly, Fig. 3 is for being described as follows showing for process referring to Fig. 3 It is intended to: in order to provide a user the service to ground of going sightseeing, is constructed using the context aware data on the ground of going sightseeing for being located at each area Sightseeing ontology and inference.

In Fig. 3, " Place " indicates the area name to be gone sightseeing, in other words, for example, such as the especially autonomous road in Jizhou (Jejudo), the special city in Soul (Seoul), crop field wide area city (Daejon), wide area city, Pusan (Busan), South Chungchong, loyalty are clear Bei Dao etc. is the same, is made of area name.

In addition, " Attraction " indicates the sightseeing ground title in each department, in other words, for example, by such as state-run natural science Shop (National museum), Jing Fuguang (Palace), Hai Yuntai (Beach), chicken Longshan, shopping center (Shopping ) etc. mall the sight name as is constituted.

At the same time, " Resource " indicates building, remains, traces and seashore etc. in each sightseeing ground, in other words, For example, by the historical building (Old in the palace (Ancient remains) of scape good fortune in utero or the place as folk customs village Building), the seashore of Hai Yuntai or Jizhou Niu Dao, restaurant (Restaurant) in department store etc. are constituted.

Especially, " Activity " indicates the practical action that user is implemented on each ground of going sightseeing, in other words, such as " Cultural activity " indicates that the cultural activity in the palace of access scape good fortune in utero, " Scenery view " are indicated in Hai Yun Mountain Niu Dao and Chinese Take of platform or Jizhou Island etc. watches the behavior of beautiful scenery with going sightseeing, and " Dining out " is indicated in department store The behavior etc. that restaurant dines out.

In addition, in embodiments of the present invention, protege_4.0.2 is used to carry out ontological construction and inference, moreover, sharp Use the buildings such as OWL (Web Ontology Language) and RDF, RDFS as the tour.owl of sightseeing ontology.

Then, referring to the outline signal that Fig. 4, Fig. 4 are the part-structure in the hierarchical structure of sightseeing ontology shown in Fig. 3 Figure.

As shown in figure 4, ontology according to an embodiment of the present invention, in order to construct sightseeing ontology, by the specified subclass in each area (subclass), and smaller group of class (class) for belonging to the area is specified, in its lower section as the specified sightseeing ground of subclass Title, and in order to illustrate in each sightseeing ground restaurant or the sightseeing ground, specify another subclass.

In addition, being provided with indicates synonym processing in order to carry out synonym or consonant retrieval to each area or sightseeing Synonymous part of speech (equivalent class) relationship.

In other words, it when user inputs the keyword to ground of going sightseeing, should be retrieved to similar word or synonym Request the case where retrieval, is arranged synonymous part of speech (equivalent class) relationship, mentioned always to user in the form of abbreviation For accurate information retrieval result.

Specifically, as shown in figure 4, as specified " Tour " for indicating sightseeing ground of highest class, on ground of going sightseeing in the following, will Each areas such as the especially autonomous road in crop field wide area city, South Chungchong, the special city in Soul, wide area city, Pusan, Jizhou are appointed as subclass, It is made of below each area the subclass for constituting the area.

In other words, it is made of below crop field wide area city East, West, middle area, scholar city, great virtue area, it is special in Jizhou In the case where autonomous road, subclass is constituted by Jizhou city and the city Xi Guipu, in the case where South Chungchong, constitutes public state and Fuyu County Deng subclass, subclass below specifies detailed address value, for example, below will in the case where the wide area city of crop field in middle area Ginkgo hole, the hole Sha Ting etc. are configured to subclass, in the case where West, by Chang'an road and the hole Wan Nian etc., in the especially autonomous road in Jizhou In the case where, in Jizhou city in the following, constituting the subclass in three 2 holes of apprentice, Jiu Zuoyi, Chinese hole etc., and below the city Xi Guipu, by Cheng Shanyi, pacify face etc. is constituted.

At the same time, subclass below is made of ground title of going sightseeing, in other words, in the case where the wide area city of crop field, Subclass is made of the middle hole Qu Shating O-World etc., and in the case where the especially autonomous road of Jizhou, by especially autonomous road Jizhou, Jizhou Lofty or bottomless cave of Jiu Zuo city, city etc. constitutes subclass relation.

It then, is that the attribute value for constituting the data value of each class is set in order to carry out ontology inference referring to Fig. 5, Fig. 5 The schematic diagram being illustrated with the process of relationship.

As shown in figure 5, constituting each regional (Place) of sightseeing ontology and the relationship on ground (Attraction) of going sightseeing has The relation on attributes of " hasAttraction " and " IsAttractionOf ", such as palace building gone sightseeing in ground and sightseeing ground Or " Resource " as remains, seashore, mountain etc. has the relation on attributes of " hasResource " and " IsResourceOf ", And user go sightseeing go sight-seeing the palace scape Fu Gong behavior, extra large holder seashore take a walk behavior, chicken Longshan mountain-climbing behavior And to the behavior etc. that department store is done shopping be " Activity ", with the relationship of " Resource " have " hasActivity " and The relation on attributes of " IsActivityOf ".

Here, there is also the feelings with multiple attributes and relationship between the data for constituting each class for each relation on attributes Condition.

It in other words, is the schematic diagram for indicating to constitute the multiple attributes relationship between the data of each class referring to Fig. 6, Fig. 6.

As shown in fig. 6, for example " Soul (Seoul) " area and in Soul with inclusion relation " Zhong Lu (Jongro) " Area has the relation on attributes of " hasSubclass " and " IsSuperclassOf ", and includes " Jing Fugong in the area " Zhong Lu " (Kyongbokgung) " and two areas " hole Ren Si (Insadong) ", to have subclass relation simultaneously.

In addition, " Zhong Lu " area and go sightseeing ground " Jing Fugong " relationship have " hasAttraction " and While the relationship of " isAttractionOf ", the area " Zhong Lu " and ground " hole Ren Si " of going sightseeing also have " hasAttraction " The relationship of " isAttractionOf ".

Therefore, the relationship is referred to as multiple attributes relationship, in each sightseeing ground, has multiple sights in identical area Light when, be considered as with the multiple attributes relationship.

Here, ground " Jing Fugong " and " hole Ren Si " of going sightseeing belongs to identical area, and therefore, the category with " atSamePlace " Sexual intercourse, and going sightseeing in " Jing Fugong ", because having the palace for tourist's visit and garden in sightseeing ground, have Multiple Resource relationship, therefore, it is known that " Jing Fugong " is respectively provided with two to " palace (Palace) " and " gardens (Garden) " The relation on attributes of secondary " hasResource " and " isResourceOf ".

It in other words, is the generalized schematic that the process of ontology inference is carried out using protege referring to Fig. 7, Fig. 7, wherein Fig. 7 a indicates the subclass relation in each area, and Fig. 7 b indicates to constitute the subclass relation on each sightseeing ground of Datian area.

At the same time, nearest user is carried out in mobile environment using the various application programs of smart phone various useful Information retrieval, as using the various application programs one of convenience, GPS function can be exemplified as.

Therefore, in the present invention, when carrying out retrieval request in mobile environment, the characteristics of using as mobile environment in The advantages of a kind of usable GPS, provides the map retrieval for utilizing GPS module, to provide various search results together.

Specifically, because can by GPS module provide user current location, by the function with longitude with The present position values of latitude determination user, and the ground locating for active user can be retrieved using the current location of the user of measurement The sightseeing that can be gone sightseeing in point ground information, in addition, for example, can provide currently using Map Services such as Google Maps (Google map) The map of position and the ground of going sightseeing retrieved.

Then, the institute of the attribute and relationship and the relationship being arranged between each class on each go sightseeing ground and affiliated ground of going sightseeing is connected After stating ontology inference process, examined to using the building of context aware data using the information of the semantic data method for digging of mobile environment The process of the sightseeing ontology of Suo Suoxu is described in detail.

In other words, in the embodiment of the present invention that will illustrate below, in order to construct sightseeing ontology, using passing through door Website relevant to going sightseeing including the website of family be collected into relevant information architecture sightseeing database of going sightseeing, at this point, Data base manipulation MS SQL Server 2000personal version has the address value on each ground of going sightseeing, and constitutes test_ Db database and tour_db table.

Specifically, Fig. 8 and Fig. 9 are the sightseeing database and sightseeing number for respectively indicating the building referring to Fig. 8 and Fig. 9 According to the schematic diagram of the part of library table.

As can be seen from figures 8 and 9, such as tour_db is by id, add_1, add_2, add_3, add_4, add_5, tour_ The domains such as site, syn, lat, long (field) are constituted.

Here, id is appointed as major key (primary key), such as in the case where the wide area city of crop field, add_1 is by crop field Wide area city is constituted, and add_2 is made of middle area, West, East, scholar city etc., and add_3 is by the hole Sha Ting, Wan Niandong, ginkgo hole, length Pacify road etc. to constitute, add_4 specifies the root park road of crop field wide area Shizhong District anvil cavern, and add_5 specifies South Chungchong public affairs state City, the temple Dong He, the face Fan Pu 327 kinds of 1 tunnel ground.

In addition, in tour_site, such as specify as being located at Jizhou especially autonomous road Jizhou city three 2 roads Dong Guande on foot No. 19 sight moral pavilions the same sightseeing ground title, and syn is when user retrieve with going sightseeing, in order to carry out synonym and set The domain setting the processing of consonant word and specifying.

At the same time, lat and long specifies the latitude and longitude in each area, here, latitude and longitude are to be, for example, Setting user location value and it is specified, the user location value is used to show map using Google Maps (google map).

Therefore, using the sightseeing constructed by the process database sharing go sightseeing ontology, in the present embodiment, In order to carry out ontological construction and ontology inference using protege_4.0.2, and utilize OWL (network ontology language, Web Ontology Language) and the tour.owl of the buildings as sightseeing ontology such as RDF, RDFS.

Specifically, the concept of tour-s ites is set as highest class, as under it to construct ground ontology of going sightseeing The specified domain add_1 (field) being present in tour_db table of layer class, such as by crop field wide area city, Jizhou especially autonomous road, loyalty Clear Nandao, wide area city, Pusan etc. are appointed as add_1 class.

In other words, referring to Fig.1 shown in 0, Figure 10 is the schematic diagram of the sightseeing body construction of the embodiment of the present invention.

It as shown in Figure 10, include crop field wide area city, wide area city, Pusan, Ji Zhoute in the upper class of tour_sites ontology The classes such as not autonomous road, South Chungchong, in its lower layer, the add_2 class that will be present in tour_db is appointed as indicating each area The subclass of upper class, such as area, East, West, great virtue are gone in crop field wide area city is specified below, and it is especially autonomous in Jizhou In the case where road, Jizhou city, the city Xi Guipu etc. are appointed as lower layer's class.

In other words, 1, Figure 11 is the schematic diagram for indicating to be unfolded the body construction of the add_2class referring to Fig.1.

Then, add_3 is specified in tour_db table as subclass, and in the add_3 class, such as in the city of crop field Area specifies the hole Sha Ting, anvil cavern, ginkgo hole, and in the case where the especially autonomous road of Jizhou, refer to below the especially autonomous road in Jizhou Determine Jizhou city, Jizhou city be specified below two on foot 1 holes, three on foot 2 holes, 1100 tunnels, Chao Tianyi, Jiu Zuoyi, Ya Yue city, seashore hole, Member of Imperial Academy city, dragon's pool road, 1 hole of dragon's pool, the face Niu Dao, writing brush capital face, arboretum road, 3 hole of dragon's pool, three Yang3Dong, 2 holes of apprentice etc..

In other words, 2, Figure 12 is the body construction that the add_3class of sightseeing ontology shown in Fig. 10 is unfolded referring to Fig.1 Schematic diagram.

Specifically, indicating to belong to Jizhou especially the Nan Yuan city in the city Xi Guipu in autonomous road area and conduct in Figure 12 The East Road Zou of subclass below, great Jing city and the West Road Ma Luolu, Zou 3000 tunnel, Xia as subclass below hold a memorial ceremony for The structures such as hole and 70 No. 214 tunnels of li as subclass below are dried in seashore road, east.

Then, as its lower layer's class, such as below crop field wide area city, the hole Zhong Qu, Sha Ting husky pavilion park road is specified For the add_4 class of tour_db table, and in its lower layer's class, i.e., under crop field wide area city, the hole Zhong Qu, Sha Ting, husky pavilion park road O-World is appointed as tour_site class by face.

In addition, in the case where the tour_site class as O-World, in order to carry out synonym or consonant processing, in order to It is specifiedSynonym relationship, synonymous part of speech (equivalent class) is set.

At the same time, 3, Figure 13 is the body construction that the add_4class of sightseeing ontology shown in Fig. 10 is unfolded referring to Fig.1 Schematic diagram, indicate the add_4 and tour_s of palace The South Pool in prefecture Fuyu County, South Chungchong Fuyu County mountain Yi Ling in tumulus, the southeast etc. The relationship of ite.

Especially, 4, Figure 14 is to indicate to utilize OWL (Web Ontology using protege_4.0.2 referring to Fig.1 Language), the partial schematic diagram of the tour.owl ontology source code of RDF and RDFS building.

Therefore, as described above, being utilized using sightseeing data as context data, so that utilizing context aware data The inference of sightseeing ontology and building may.

In other words, referring to Fig.1 shown in 5, Figure 15 is according to the embodiment of the present invention for providing based on context aware User pays close attention to the integrally-built generalized schematic of the context aware ontology implementation method of information.

Specifically, context aware ontology implementation method according to an embodiment of the present invention includes series of steps: receiving first Collect and information database (S151 step) is paid close attention to the information architecture user of user's region-of-interest, execution setting is stored in constructed User pay close attention to information database each data between relationship and attribute ontology inference after (S152 step), according to ontology The relationship and attribute of set each data in inference step, construct context aware ontology database (S153 step).

Here, the building user pays close attention to information database and constructs the specific of ontology database by carrying out ontology inference Process can refer to Fig. 2 to Figure 14, can with it is described above building sightseeing ontology process together with realize, here, should notice for Simplified explanation, is omitted and illustrates to each step.

Then, using the context data library constructed as described above to based on discovery and similar or identical frequent of user query The semantic data of the FP-growth algorithm of mode is excavated (Semantic Data Mining) method and is illustrated.

In other words, semantic data, which is excavated, refers to the structure possibility in order to enable the logic to various data, and to user Inquiry provides useful and fast and accurately information, finds the process of hiding mode, below illustrated implementation of the invention In example, in order to effectively transmit suitable and useful letter to user using the existing FP-growth algorithm based on FP-Tree Breath, it is right using the data mining process and context aware data (Context Awareness Data) of similar query pattern to illustrate The inquiry of user provides suitable service and executes a series of processes that semantic data excavates (Semantic Data Mining).

Here, to execute data mining process, needing following several definition as example proposing in [Gu 2010] (referring to bibliography 15).

[defining 1] (XML Query Pattern Tree)

XML query scheme-tree is the subtree (rooted tree) with root, by XQPT=<X.V, X.E>composition.

Here, X.V is vertex set (vertex set), X.E is side collection (edge set), and root is with Root (XQPT) table Show, each side using as vertex (v1, v2) expression, v1 be v2 parent (parent), each vertex v have have *, //, TagSet } value label (label), tagSet indicates to be present in all elements (element) and attribute of DTD (attribute) name set.

[defining 2] (XML Query Rooted Subtree)

When provide query pattern tree XQPT=<X.V, X.E>when, the XML query subtree for having root of the subtree as XQPT XQRST=<X.V', X.E'>, meeting such as Root (XQRST)=Root (XQPT) andEqually Inclusion relation when, become the subtree of XQPT.

In other words, the structure for searching the useful subtree to take place frequently is known as XML query subtree, and searches based on this similar Structure, at this point, the tree found is referred to as XML query scheme-tree.

Specifically, 6, Figure 16 a and 16b is to respectively indicate XML query scheme-tree (Query pattern referring to Fig.1 Tree) the exemplary diagram of the XML query subtree (Query Rooted subtree) of He Yougen.

As shown in figure 16, query set (query set) is made as query pattern tree and is obtained after being stored to it XQPT_DB={ XQPT as query pattern tree database₁,...XQPT_n, and query pattern excavates (query pattern It mining) is to search to meet all with the root that takes place frequently of minimum support (minimum support) in database XQPT_DB Subtree.

In other words, in XQP_DB, all situations including the subtree XQRST with root are with indicating degree of taking place frequently FRQ (XQRST) indicates, here, to acquire support (support) value of XQRST, using following mathematical expression SUP (XQRST)= FRQ (XQRST)/| XQP_DB |, the quantity that the degree of taking place frequently of XQRST is divided into all database is calculated.

In addition, if XQRST meets SUP (XQRST) >=σ condition, meaning for the σ (σ > 0) with certain value XQRST takes place frequently in database XQP_DB.

It is the structure for respectively indicating three kinds of query pattern trees and the subtree that takes place frequently with root for example, referring to Figure 17, Figure 17 Schematic diagram.

In Figure 17, XQRST occurs in XQPT1 and XQPT2, therefore, the subtree (Rooted Subtree) with root Degree of taking place frequently be FRQ (XQRST)=2, and SUP (XQRST)=2/3.

As described above, the process for excavating the query pattern that takes place frequently is to search similar query pattern to improve search efficiency Process, in the process, for find query pattern whether be it is similar or identical, need to inquire shown in following [defining 3] Pattern matching process.

[defining 3] (XML Query Pattern Tree Matching)

When having query pattern tree XQPT and subtree XQRST with root, if meeting following condition, XQRST is contained in In XQPT.

1, the root node (root node) of XQRST and XQPT shares identical label (label).

If 2, some node w ∈ XQRST of XML tree structure is consistent with v ∈ XQPT, there will be w.label≤v.label Relationship, and each subtree of w have be contained in XQPT subtree relationship.

In other words, the XQRST of the query pattern structure proposed as user between XQPT structure, each node it Between relationship there are when inclusion relation as described above, then may be defined as XQRST to be contained among XQPT, and the inquiry mould of user Formula tree becomes frequent pattern tree (fp tree).

Hereinafter it is assumed that illustrating the process for searching the answer to inquiry after user performs following inquiry.

Q1: please find out type (genre) as movement (action), make in the film that annual (year) is 1994, main Angle (actor) is the title (title) of the film of " Jean Reno ".

For the answer for finding the inquiry, the data as genre, year, actor need to be indicated using based on the inquiry With the path of the value to be searched and they.

In other words, with //year=" 1994 " and//genre=" action " and//actor=" Jean Reno " It indicates inquiry, and is used using the path as an affairs, excavated by the semantic data based on FP-growt algorithm Journey executes the operation that similar mode is searched using the query pattern of user.

Therefore, using basic definition as described above, the semantic query mode excavation to XML data is by following process Execute: building frequent pattern tree (fp tree) (FP-Tree) simultaneously searches the XML query mode configuration to take place frequently, thus the inquiry of discovery and user Similar or consistent useful subtree (with reference to bibliography 16 and bibliography 17).

Here, in order to used in the present invention compared with as existing by the process model mining query pattern that takes place frequently Search take place frequently structure algorithm Apriori algorithm in terms of time and space more effective FP-growth algorithm, institute as above It states, in order to execute the data mining to the tree schema that takes place frequently using FP-growth algorithm, needs following building frequent pattern tree (fp tree) (FP- Tree process (referring to bibliography 18)).

Specifically, the process of building frequent pattern tree (fp tree) (FP-Tree) is, firstly, XML query mode is set to a thing It is engaged in, and investigates the degree of taking place frequently of each project (item) by preliminary sweep in the database being made of the affairs.

Here, degree of taking place frequently indicates the support angle value of element (element) project set (items sets).

The second, the project in each affairs is arranged by degree of taking place frequently descending.

Third, in the project arranged by degree of taking place frequently descending, remove the minimum support (minimum that specifies than user Support) few project.

4th, in the step by the project not taken place frequently from transaction database remove after, utilize remaining project It constructs frequent pattern tree (fp tree) (FP-Tree).

Here, frequent pattern tree (fp tree) (FP-Tree) is prefix (prefix) tree, each path (path) is shared using in tree The affairs of identical prefix (prefix) indicate, and each node indicates a project.

It therefore, can be by repeating process building frequent pattern tree (fp tree) (FP-Tree).

In other words, 8, Figure 18 is in the building frequent pattern tree (fp tree) (FP-Tree) executed by the process referring to Fig.1 It shares prefix (prefix) process in the process and indicates the generalized schematic of the process of each affairs.

In Figure 18, frequent pattern tree (fp tree) (FP-Tree) is to indicate database in a compressed format, in header (Header) table In be connected with each project and the pointer (pointer) as node link.

As described above, utilizing the FP- for searching XML query frequent mode by building frequent pattern tree (fp tree) (FP-Tree) The data mining process of growth algorithm is realized by executing following process: selection meets the project of minimum support, and goes Except the project for being less than minimum support.

In addition, 9, Figure 19 is to indicate during executing FP-growth algorithm and carry out data mining referring to Fig.1, structure Build the schematic diagram of the process of condition (conditional) frequent pattern tree (fp tree) (FP-Tree) as Arbitrary Term purpose m.

As shown in figure 19, the process of building condition tree and search criterion mode is repeated to each project.

At the same time, referring to Figure 20, Figure 20 be expression as the process for executing the search criterion mode repeatedly as a result, The schematic diagram of condition pattern obtained is arranged by table.

Then, illustrate the example that the data digging method is applicable in when user inquires.

First, it is assumed that having carried out XML query based on following XQuery Syntax.

[inquiry]: in the book written by the people for possessing the name of " Kim ", retrieval title, writer and price.

In addition, Figure 21 is the exemplary diagram that the Korean inquiry for executing user is made as XML form referring to Figure 21.

Here, lastname=" Kim " indicate the book written by the people for being called " Kim " in retrieve its book title, The content of author, price, resultPattern indicate the mode normal form (schemapattern) to query result, Predicates indicates that the filter condition required when executing inquiry, documents indicate relevant XML file.

It at the same time, is to indicate as the Book based on XML document made by the Xquery referring to Figure 22, Figure 22 The schematic diagram of DTD Tree.

In Figure 22, Book DTD Tree indicates the form for the database being made of affairs.

Specifically, in order to execute the excavation found to the data of the query pattern that takes place frequently of inquiry from the user Journey, firstly, building will become the database of an affairs by each query pattern tree caused by the inquiry of user, to each Query pattern tree is made whether to be checked and stored for useful query pattern, thus the database that building is made of affairs.

Then, by searching the root tree having in the database for being stored with query pattern tree, and by searching for this The process of similar mode executes data mining.

In other words, as described above, generating frequent pattern tree (fp tree) (FP-Tree) based on the affairs being made of each query pattern The structure of form, and the side FP-Growth utilized for searching frequent pattern tree (fp tree) is executed based on the FP-Tree generated as described above The data mining process of method is taken place frequently tree by the process model mining, is taken place frequently subtree (maximal frequent to search maximum subtree)。

Here, the inclusion relation of the tree required during described to lookup is defined as follows:

[defining 4] (Tree Subsumption)

When there are two the S and S as subtree collection₁When, it is S=S₁∪{t₁, t₂,...,t_nRelationship.

Here, ift_j<t_i, then S₁It is contained in S, is expressed as S₁<S。

[defining 5] (Maximal Frequent Subtree)

If the collection of subtree is combined into the subtree and be not the relationship for including, the subtree by other any subtrees that take place frequently of taking place frequently Collection is combined into the subtree (Maximal Frequent Subtree) most to take place frequently, in other words, exists Relationship.

As described above, in an embodiment of the present invention, in order to improve the efficiency to user query, using based on existing The FP-growth method of association rule algorithm, and the extension as FP-growth method look into the similar of XML data searching On the basis of the method for inquiry mode, increase that differentiation takes place frequently subtree and non-take place frequently and goes the mode unless the subtree that takes place frequently at subtree.

In addition, by searching max model subtree (maximal pattern subtree) using FP-tree algorithm, it is right Query pattern from the user carries out analysis and by being compared with max model subtree, divides user query mode Class, it is whether consistent with max model subtree by the Classification Count user query mode, by executing the investigation ratio repeatedly Compared with the subtree and max model subtree of user query, thereby executing the mining process for searching max model subtree.

Then, to the data mining process progress process searched using context data to the scheme-tree of user query.

In other words, in an embodiment of the present invention, using context data (Context Data), pass through FP-growth Scheme-tree data mining (the Pattern Tree Data Mining) method of algorithm executes semantic data and excavates.

At this point, context data is to collect sightseeing number with the internet site for possessing each sightseeing information using portal website According to, and this is gone sightseeing using the data building sightseeing database being collected into, and using widely used Protege_4.0.2 building Body.

In addition, in sightseeing ontology, using the OWL (Web Ontology Language) as Ontology Language by RDF and RDFS is indicated, to generate the ontology database for being referred to as tour.owl, will belong to this area on a small scale based on each area Area is appointed as the subclass under it, moreover, by going sightseeing being appointed as another subclass under the subtree of multilayer.

At the same time, subclass, superclass, synonymous part of speech using going sightseeing are specified with the relationship of each department and attribute Classes such as (equivalent class), and in each area, in the relationship for belonging to a regional multiple sightseeing ground, specify class It is similar to the attribute of at the same place.

Especially, when user carries out retrieval request, in order to carry out synonym and the processing of consonant word, synonym is specified The famous restaurant on belonging each ground of going sightseeing or the content comprising the explanation to the ground of going sightseeing are appointed as attribute by class, and are passed through Position Latitude attribute is serviced using Google Maps (Google map) each contextual information for going sightseeing ground being supplied to user.

Therefore, the ontology database of the building has the hierarchical structure using RDF and OWL, accordingly, in the present invention, In order to execute data mining process using sightseeing ontology, XML data method for digging is excavated into mode using for master data, and And in order to search the query pattern of incidence relation to user, looked into be searched using existing FP-growth algorithm with user Same or similar sightseeing ground is ask, the scheme-tree data mining side for searching the same or similar mode for being stored in ontology is applicable in Method.

Specifically, tour is appointed as root, using the sightseeing for constituting sightseeing database information is by each regional structure As next subclass relation, in its lower setting relationship, to constitute another subclass, so that each area and sightseeing will be constituted The hierarchical structure on ground is appointed as an affairs.

In addition, when user with retrieving the sightseeing of given area when because to the sightseeing for constituting inquiry from the user Address value, it may have such as the hierarchical structure the structure being stored in ontology database, therefore, by searching for user The process of the structure of inquiry affairs similar or consistent with the affairs for the hierarchical structure being stored in ontology database executes number According to mining process.

At the same time, referring to Figure 23, Figure 23 is the project set that takes place frequently that each affairs are indicated in sightseeing ontology database The schematic diagram of (transaction frequent itemset)

In Figure 23, by the structure of each hierarchical structure for constituting affairs with going sightseeing and specific one ground of going sightseeing being constituted Table is constituted as each affairs, and indicates to constitute the content on each sightseeing ground of affairs id and the affairs to take place frequently.

Especially, referring to the sightseeing that Figure 24, Figure 24 are by utilizing the structure that takes place frequently occurred by each area as described above The partial schematic schematic diagram of the FP-tree structure of ontology database.

In Figure 24, project (item) is constituted using each area name in left side, and each affairs will be constituted on right side Project used as a node, to indicate the link of node.

In addition, numerically constituting the degree of taking place frequently of the project of each node beside it.

In other words, such as in sightseeing ontology database, in the case where the especially autonomous road of Jizhou, because of the feelings to take place frequently Condition occurs 5 times, so indicated with 5, and in the case where the city Xi Guipu, because occurring 3 times, indicates to have and cries 3 to take place frequently Degree, to other each area names, continues to mark as described above, to construct FP-tree structure.

At the same time, the condition FP-tree to some specific project can be constructed based on the FP-tree.

It in other words, is the whole world (Global) FP-tree referring to Figure 25, Figure 25 a, and Figure 25 b is in area of going sightseeing The schematic diagram for the case where condition (Conditional) FP-tree is constructed on the basis of Ma Luolu.

As shown in figure 25, FP-growth algorithm execution repeatedly generates some specific project based on constructed FP-tree The process of condition FP-tree, and searched and repeating process of the generation to the condition FP-tree of each project The process of FP-tree structure is data mining process.

Specifically, for example when user issues the inquiry of retrieval " the long Mount Taishan mastery woods in crop field wide area city West Chang'an road " When, in order to search the inquiry similar with the inquiry of user, the process of investigation sightseeing ontology database is executed, when presence and the use When the identical structure of the inquiry at family, service can be quickly and accurately executed to user, but when, it is searching only and " crop field wide area city west When the consistent structure of head of district An Lu ", because being similar but not being consistent structure, offer is not very accurate but similar knot The result of structure.

It at the same time, is the part source generation for indicating to execute the FP-growth algorithm of process as described above referring to Figure 26, Figure 26 Code schematic diagram (refers to bibliography 19).

Then, the process that data mining is executed using FP-growth algorithm as described above is described in more detail.

Firstly, executing the preprocessing process encoded to data in order to which ontology database is suitable for data mining.

For this purpose, the data in belonging each area are transformed to code appropriate and are applicable in FP-growth calculation by the present inventor Method observes the process for generating which type of mode with this.

Here, FP-growth algorithm is also suitable Minimum support4 as the Apriori of applicable associations rule (minimum confidence) and minimum support (minimum support) is therefore observed applicable as described above critical Be worth (threshold) when applicable data excavate the result is that.

Specifically, referring to Figure 27, Figure 27 be indicate to be applicable in minimum support in FP-growth algorithm be 0.4 and Carry out the schematic diagram of the result of data mining.

As shown in figure 27, the quantity for the affairs being stored in database profession is 99, and maximum memory usage amount is 0.96mb, frequency Numerous project set (frequent itemset) generates 27, and executing the time consumed by data mining is 15ms.

In addition, can be observed when executing five steps of L1 to L5, with minimum support (minimum support) value While being applicable in support angle value on the basis of 0.4 respectively, mode is generated by each step, here, the value in each mode is meaning To the value that is deformed of value in the area in database.

Therefore, in Figure 27, it can be seen that the frequent mode on the ground of going sightseeing in the city Xi Guipu for being located at the especially autonomous road in Jizhou It occurs most frequently.

More than, it is illustrated, is connect using the process that context data executes data mining to based on FP-growth algorithm , realize that the overall process of information retrieval is described in detail by executing semantic data mining process in mobile environment.

In other words, the present invention is to accurately compactly provide useful information of the user needed for current location, including each Including the setting of class relationship and attribute setting in a ground section, set to show the relationship of word of the word or identical meanings that indicate similar Set and using sightseeing ontology, and in embodiment as described above, for sightseeing regional needed for the accurate transmission user information And construct sightseeing ground ontology.

In addition, dividing different regions that will be ontology, example to the Informational Expression on ground of going sightseeing using hierarchical structure in ground ontology of going sightseeing Such as when user issues " sightseeing of the retrieval positioned at crop field city, area, scholar city in the state of being currently located at crop field city, area, scholar city When the inquiry on ground ", it is contained in the scholar city area in crop field wide area city by ontology retrieval, and is included in scholar city to user Sightseeing show as search result, at the same time, for example, as Google Maps (Google map), by being built in The map function of mobile device provides the map on ground of going sightseeing required for current location and user, moreover, a series of by integrating Process the retrieval service to the ground of going sightseeing for being located at the area nearest with user location needed for user is provided.

Especially, it is dug as described above, executing data to the data in sightseeing ontology database using FP-growth algorithm Pick process, provides a user useful information, in the following, to the semantic information retrieval for being applicable in the semantic data mining algorithm and executing Process is described in detail.

Firstly, referring to Figure 28, Figure 28 be indicate using the method for mode matching based on semantic data mining algorithm based on The integrally-built generalized schematic of mobile information retrieval system.

In system as shown in figure 28, operation performed by modules and overall flow are described as follows:

The first, user is in current location execution information retrieval request.

In other words, when user initiates retrieval request using mobile device, WiFi, 3G, 4G, GPS, Sensor etc. are utilized Information to user current location is transmitted together with user search requirement item, to collect the contextual information of user (Context Information)。

The second, context data library (context database) is constructed.

Here, in the present embodiment, in order to collect ground information of going sightseeing, collecting each the Internet portal website and each area The sightseeing of introduction ground information, in order to each area each sightseeing and the sightseeing the characteristics of and address and periphery restaurant and paddy The setting of map (Google map) is sung, the longitude and latitude value in each area are collected.

In addition, being stored in situation number after extractive process and preprocessing process by the data being collected into as described above According to library (context database).

Third executes ontology inference process to the data for being stored in context data library by the process, and each number is arranged According to the attribute and relationship of value, to construct ontology database.

Here, in ontology inference process, in order to which the subclass (subclass) in each area, superclass (superclass) close System and synonym or consonant word are handled, and synonymous part of speech (equivalent class) relationship, each area and sightseeing ground is arranged Between " is attraction of " relation on attributes, performance belong to it is identical area in go sightseeing ground relationship " at the same The attributes such as place of ".

4th, it in order to improve recall precision using the database for being stored in ontology, executes and searches and inquiry from the user The semantic data mining process of similar or consistent query pattern tree.

Here, being using existing using the process of the requested retrieval information of user and contextual information search pattern tree FP-Growth algorithm executes data mining process.

5th, pass through context aware Information Mobile Service interface (Context Awareness Mobile Service Interface it) is provided by executing result acquired in the process as the result to information required by user.

Therefore, by the process, it can be achieved that realizing information by executing semantic data mining process in mobile environment The general frame of retrieval is then described in detail in mobile environment and is provided needed for user using the retrieval of semantic data method for digging The overall process of information.

It in other words, is performed when indicating to receive the retrieval request to ground of going sightseeing from user referring to Figure 29, Figure 29 A series of generalized schematic for the treatment of processes.

Specifically, as shown in figure 29, if first, user passes through service request module (Service Request Module) the information retrieval to ground of going sightseeing to be retrieved in person of request, then come from using the mobile search network analysis of ontology The retrieval of user requires item and constitutes query pattern.

The second, the current information that user is collected by situation processing module (Context Process Module), changes sentence It talks about, as described above, including the information content by the acquisition user query such as WiFi, 3G, 4G, GPS, sensor (Sensor) , user requires the current location under the situation of retrieval and the requirement item of user.

Third, sightseeing content (Tour Contents) DB can be referred to as the context data library (Context Database), be the sightseeing to each department including portal introduce website sightseeing information mobile phone and construct Sightseeing ground database.

Here, go sightseeing ground database in, pass through execute preprocessing process refine data and be stored in relevant database it Afterwards, by executing ontology inference process building sightseeing ground ontology database.

Therefore, as described above, in order to be looked into using the query pattern being stored in the context data library for being stored as bulk form Look for the user requested consistent mode of user query mode, be made of by retrieving the hierarchical structure relationship to ground of going sightseeing Sightseeing ontology database finds query pattern similar or identical, at this point, as described above, by using based on FP-growth The semantic data mining algorithm of algorithm searches query pattern similar or identical.

4th, the letter that user requests retrieval is retrieved in ontology database by retrieval module (Retrieval Module) The query pattern of breath, and search result needed for providing a user user together with GPS information.

Then, illustrate the data mining process for searching scheme-tree similar or identical.

Firstly, collecting information relevant to going sightseeing using internet sites etc. and constructing database, and utilize the number It is referred to as to search the pattern database (Pattern of user query mode when user inquires according to the ontology database of building Database), various modes are stored with and in the pattern database (Pattern Database), therefore, when user holds When row inquiry, execute with the comparison procedure of present mode, thus by searching for consistent mode or similar mode and inconsistent Mode process and execute data mining.

Therefore, when finding mode identical or similar with user query, user institute can be quickly and accurately provided to user The useful information needed.

In other words, referring to Figure 30, Figure 30 is the schematic diagram being illustrated to following process: receiving retrieval from user and asks When asking, the requirement item by analyzing user analyzes the query pattern of user, and by comparing consistent with the query pattern of analysis Mode and be stored in searching system ontology database mode configuration, retrieve mode configuration similar or identical.

Here, pattern database refers to sightseeing ontology database in Figure 30.

In other words, if user initiates the inquiry request to ground of going sightseeing in current location, passing through preprocessing process In the sightseeing ontology database stored with ontology inference process, the query pattern of investigation active user's request and existing deposit Be stored in the query pattern of pattern database it is consistent whether, show search result in the case where consistent and similar, and different Again search pattern in the case where cause.

Specifically, if user ask my the desired retrieval for going sightseeing ground in the retrieval window of searching system It asks, to obtain retrieval request from the user, then user is generated by analytic process of the execution to the query pattern of user and looked into Inquiry mode.

Then, check whether user query mode is new mode, when user mode is new model, investigation whether with it is existing Have that mode is consistent, if there is consistent mode, then will be stored in the consistent mould of pattern database (pattern database) Formula is provided as search result.

If can not investigate new mode it is consistent with the mode that is stored in pattern database whether, again described in execution Process provides similar mode as search result when there is similar mode, and the new mode of user query is with new Mode is stored in pattern database.

It therefore, will be with user as described above, executing a series of processes repeatedly according to an embodiment of the invention, passing through The identical or similar query pattern of inquiry provided as search result, therefore, it can be achieved that from using sightseeing data as situation number In the sightseeing ontology constructed according to utilization, user is searched using the semantic data mining algorithm based on FP-growth algorithm and is paid close attention to The search method of information, in addition, can provide the user based on context aware accordingly pays close attention to information recommendation service.

It in other words, is using semantic data excavation according to the embodiment of the present invention based on situation referring to Figure 31, Figure 31 The user of perception pays close attention to the integrally-built generalized schematic of information retrieval service providing method.

Specifically, as shown in figure 31, it is according to an embodiment of the present invention using semantic data excavate based on context aware User pay close attention to information retrieval service providing method, firstly, if user using oneself include smart phone or tablet computer Mobile device inside sends the retrieval request to concern information in current location, then utilizes the WiFi of terminating machine, mobile communication Net (3G, 4G etc.), GPS and terminating machine belonging to being set to various sensors and meanwhile receive the retrieval request of the user and to institute The information of the current location of user is stated, and is collected as the contextual information of the user (Context Information) (S311 step).

Then, extractive process and preprocessing process are executed to collected information and constructs context data library (context Database) (S312 step), and each data being stored in constructed context data library are arranged by ontology inference The attribute and relationship of each data value, to construct ontology database (S313 step).

Then, using the ontology database constructed by the process, it is based on the FP-Growth algorithm, executes lookup The semantic data excavation of query pattern similar or consistent with the inquiry that user requires handles (S314 step), and will pass through semanteme The result that data mining processing step obtains is shown in the movement of user as the search result of the retrieval request required user Terminal (S315 step).

Here, the more specific content of the processing executed in each step can refer to existing technical literature and Figure 16 To simplify explanation, the detailed description to each step is therefore omitted herein to content illustrated by Figure 30.

Then, needed for being described in detail to realize the mobile message searching system as described above using context aware ontology Environment and its realize process.

In other words, in the described embodiment, the present inventor is to collect sightseeing data, utilizes the introduction including portal The website on the ground of going sightseeing in each area collects ground information data of going sightseeing, and executes the pretreatment refined to collected information Process, such as in order to construct relevant database in the operating system environments such as Windows, utilize MS SQL Server2000Personal version constructs database.

In addition, respectively constituted in database id, add_1, add_2, add_3, add_4, add_5, tour_site, The domain (field) of syn, lat, long and storing data, here, as add_1, add_2, add_3, add_4, adde_5 Each domain representation constitutes the address on each ground of going sightseeing, and tour_s ite storage sightseeing ground title, syn is for synonym or consonant The domain name of word processing, and lat and long respectively indicate the latitude and longitude of setting map (Google map) Shi Suoxu.

At the same time, the present inventor pass through using between each data of relational data lab setting relationship and attribute simultaneously It constructs sightseeing ontology (Ontology), at this point, executing sheet using Protege_4.0.2version to construct ontology database Body building and inference, and ontological construction tour.owl ontology is showed using OWL (Web Ontology Language) and RDF.

Especially, the present inventor executes the semantic number based on FP-growth algorithm using the ontology database of the building According to the information retrieval system for excavating and constructing mobile environment, here, for realizing in the realization environment of mobile message searching system, Such as using [Android 2014] 4.3 version of Android proposed (referring to bibliography 20) and for realizing Google Scheme (Google map) Google APIs 4.3, in order to realize mobile message searching system, be utilized HTML5, CSS3, JavaScript etc., and in order to be utilized as database and server, the Node JS for high-capacity database is utilized server。

In addition, being constituted using Jdk1.7.0_05 because Android programming is based on Java and realizing environment, moved for constructing The WebStrom_ for being for example set forth in [WebStorm 2014] is utilized in the editing machine of the information retrieval system in rotating ring border 7.0.3version (bibliography 21 is referred to).

Here, if to the Android programming as the mobile platform (Mobile Platform) for being suitable for the present embodiment Feature is illustrated, then as mobile device platform Android have a characteristic that first, Android be Google production, Disclosed open source mobile platform is suitable for the operating system based on Linux of mobile environment, user section, application program Set；The second, different from closed mobile phone environment, so that the exploitation and operation of the various application programs of restrained minimization may.

At the same time, to the detailed content of the development process using Android, those skilled in the art can be for example, by The existing technical literature of [Michalis 2012] (referring to bibliography 2) is readily appreciated that therefore, details are not described herein.

In other words, as described above, the present inventor is using the Internet portal website or introduces the sightseeing information in each area Website information architecture sightseeing database, and in affiliated sightseeing database constitute id, add_1, add_2, add_3, add_ 4, the domains such as add_5, tour_site, syn, lat, long, herein.Such as add_1 is by " crop field wide area city ", " Jizhou is especially autonomous The area names such as road ", " wide area city, Pusan ", " South Chungchong " constitute, add_2 in the case where the wide area city of crop field by " West ", Areas' title such as " East ", " middle area ", " scholar city ", " great virtue area " is constituted, and add_3 is such as located at " 10000 years of crop field wide area city West Hole title is equally specified positioned at " the ginkgo hole " of crop field wide area Shizhong District etc. in hole ", and add_4 is such as located at crop field wide area city West ten thousand " the village mountain main road " in year hole, positioned at equally specified better addresses such as " in will in " in the city the South Chungchong Gong Zhou face Ji Long, and add_5 " No. 169 " of the specified village mountain main road such as crop field wide area city, 10000 years holes, West, be located in the city the South Chungchong Gong Zhou face Ji Long will in The lot numbers such as " No. 52 ".

In addition, specified village mountain main road 169 for being for example such as located at crop field wide area city, 10000 years holes, West tour_site is " big Field culture and arts garden ", sightseeing in the city the South Chungchong Gong Zhou face Ji Long by inner No. 52 " first temple " equally ground title, Syn is for example in order to " crop field " being appointed as to " Korea Spro bar (hanbat) " or when " Jizhou " is appointed as to " indulging in sieve (tamra) " is made Synonym processing, or in order to input the consonant word as " first temple " when inputting search keyAs The consonant word of crop field culture and arts gardenDeng the processing of consonant word and specify, and lat and long The longitude and dimension values in each area are specified in order to indicate map (Google map).

At the same time, in described the embodiment of the present invention, the sightseeing database stored as described above is for ontology of going sightseeing The building of database, and in order to carry out ontology inference and ontological construction using the attribute and relationship of each data in database And Protege_4.0.2version is utilized, and construct using OWL (Web Ontology Language) and RDF The ontology database of tour.owl.

Here, the attribute and relationship of the specified data value for indicating each area of ontology inference process, in other words, specified pair The superclass (superclass) and subclass (subclass) relationship in each area and multiple inclusion relation, specify for synonym and The synonymous part of speech (equivalent class) of consonant word processing, in the case where the sightseeing ground being contained in identical area, if The relation on attributes as " at the same place of " is set, in the case where the sightseeing ground being contained in area, such as Positioned at the Jizhou especially autonomous road Jizhou city island Niu Dao Mian Niu sightseeing ground, such as " is can be set on area with going sightseeing relationship The relation on attributes of attraction of " equally.

Especially, in described the embodiment of the present invention, in order to using the ontology database constructed by the process to User provides more acurrate quick information retrieval structure, executes the semantic data based on FP-Growth algorithm and excavates (Semantic Data Mining) process.

Here, the data digging method will constitute higher level's class and junior's class and sightseeing in each area of ontology database The hierarchical structure relationship on ground, which is considered as an affairs (transaction) and is considered as, constitutes a scheme-tree, and based on storage in this way Scheme-tree look into and search by comparing the scheme-tree of user query mode and existing storage in the inquiry for receiving user The process of identical or similar scheme-tree carries out.

Therefore, can be provided by executing the data mining process of the process it is consistent with the search key that user requires or Similar search result.

Then, illustrate to be constructed by the semantic data mining process based on FP-Growth algorithm in mobile environment, Carry out the process of the data retrieval system of the retrieval to user query.

Here, in the embodiment of the present invention that will illustrate, in order to realize the information retrieval system of mobile environment and Realize that the present invention is described for searching system using programs such as HTML5, CSS3, JavaScript, but it is of the invention It is not limited to this, can be realized as needed using any other language or platform.

In other words, if programming realization system using Android (Android), setting with Android system can only be operated in It in standby, but has the following advantages that as described above, if being programmed using HTML5 and CSS3, not only may operate in Android system System, also may operate in iOS system.

In addition, the event-action (Event action) in the embodiment of the present invention that will illustrate, as button JavaScript is utilized in the processing such as function, at this point, the program for being used as editing machine is WebStrom_7.0.3version (ginseng Examine bibliography 21).

At the same time, it for the system for executing building, is utilized 4.3 version of Android (with reference to bibliography 20), and In order to realize the map (Google map) in each area, Google APIs 4.3for Google map is utilized.

Specifically, Figure 32 is for realizing the signal of the part of original code of the initial picture of retrieval window referring to Figure 32 Figure.

In other words, behind sightseeing ground needed for source code representation shown in Figure 32 is inputted in user by retrieval window, under One page shows its result with catalogue form.

In addition, Figure 33 is to be realized with script function (Script function) to index button (search referring to Figure 33 Button the source code schematic diagram of event-action (Event action)).

In other words, the button actions that source code representation as shown in figure 33 performs the following operations: user to retrieval window When mouth inputs search key and clicks the button beside it or click three buttons for being located at key frame lower end, conversion page, To show end value in picture by clicking button.

It at the same time, is to indicate that the map executed for providing map (Google map) service is pressed referring to Figure 34, Figure 34 The part of original code schematic diagram of button movement.

In other words, source code as shown in figure 34 is to show Google Maps using the event-action and function of btn-map Movement, indicate using area latitude value and longitude show belonging to each position map movement button actions.

It especially, is the structure according to the key frame of the mobile message searching system of the embodiment of the present invention referring to Figure 35, Figure 35 At exemplary diagram.

As shown in figure 35, if the retrieval window in the key frame in left side inputs area name or sightseeing ground title and clicks Index button then exports the search result in belonging affiliated area or ground of going sightseeing, in other words, for example, if defeated to retrieval window The ground zones values for entering " crop field " export sightseeing ground relevant to crop field title then as right panel.

Here, search result value will be shown if address value input is has the other values of subclass relation with crop field wide area city Show more detailed result.

In addition, as the right panel of Figure 35, in the lower end of all pages of the key frame except searching system (footer) part display multiple buttons, in other words, for example, first button is the button that back to homepage, second button For read be input to retrieval window and be worth call relative region map map button, third button is for showing Belong to restaurant's title in the area on related ground of going sightseeing or the content page button of the introduction to ground of going sightseeing.

Specifically, Figure 36 is the ground zones values and point respectively indicated to retrieval window input crop field wide area city referring to Figure 36 It composition shown by the map in affiliated area and shows according to the input value inputted to retrieval window when hitting map button and introduces phase Close the schematic diagram of the composition of the picture in the restaurant in area or area of going sightseeing.

Here, the map button is in order to be arranged map (Google map) in the explanation to the sightseeing ontology, it is sharp With the latitude and longitude being arranged in the database with the thresholding of lat and long, with, the map of relative region called as result Value.

In addition, Figure 37 is to indicate with the present position values of latitude and longitude automatic sensing user referring to Figure 37 and Figure 38, To realize that the source code for the regional map and latitude and the map of longitude for indicating that user is presently in together to user shows It is intended to, and Figure 38 is the schematic diagram for indicating the result screen realized by source code as shown in figure 37.

As shown in figure 37, it is as follows to realize that source code may be configured as map: the ground in order to realize belonging user current location Scheme and create myLocation function, and uses navigator.geolocation.getCurrentPosition letter wherein Number acquires the present position values of user.

At the same time, in the result screen shown in Figure 38, as the present position values of user, 36.30 He of latitude is utilized The value of longitude 127.36 acquires the positional value in crop field wide area city West Zhenglin hole, and shows the map of its surrounding area.

Therefore, the information of the current location of the present position values transmitting user using user is realized respectively by the process Movement and the movement realized when user inputs search key, excavated to can realize and be applicable in semantic data in mobile environment The information retrieval system of method.

Then, to the experiment knot using the practical execution information retrieval of the mobile message searching system realized by the process Fruit and corresponding performance evaluation are illustrated.

In other words, the present inventor constructs sightseeing ontology database by the process, and using in affiliated database Data and applicable data mining algorithm and execute experiment.

Here, the embodiment of the present invention being described below executes experiment, it is applicable in by each area for examining with obtaining The search result of the semantic data method for digging on the ground Suo Guanguang as a purpose, will be considered as one with relevant to each area going sightseeing Affairs simultaneously analyze frequent mode, in addition, in order to sightseeing needed for user quickly provide search result, to the standard of information retrieval Exactness is also tested.

Firstly, the experimental result for being applicable in semantic data mining algorithm is described as follows: the present inventor in order to retrieve with The relevant sightseeing ground in each area, as searching the algorithm of the user query mode that is mutually related using FP-growth algorithm, and The data for being utilized respectively ontology database and existing relevant database compare and evaluate performance.

In other words, it is executed using the semantic data mining algorithm based on FP-growth algorithm between two data of discovery Frequent mode process and execute the experimental evaluation that is compared such as time and memory usage amount, commented to carry out the experiment Valence executes the preprocessing process to sightseeing ontology data.

Specifically, the letter about each ground of going sightseeing of the first, website collected various portals and introduce sightseeing ground Breath, to construct sightseeing database；The second, it using the data being stored in sightseeing database, constructs as sightseeing ontology Tour.owl, here, the ontology is constructed using protege_4.0.2, and inference and performance are that OWL, RDF, RDFS is utilized.

Third, the parsing to sightseeing ontology database using DOM parser (Parser) execution to XML document；4th, it is It extracts the label (tag) of XML document and executes FP-growth algorithm, create map table (mapping table), and to each A label carries out numeralization, thereby executing mining process.

In addition, in the case where the data of existing relevant database, also extract sightseeing data and by with described The similar process of volume data executes preprocessing process, to create map table and carry out numeralization and execute data mining process.

It in other words, is to indicate through table to according to an embodiment of the present invention referring to Figure 39 and Figure 40, Figure 39 and Figure 40 FP-growth algorithm is applicable in volume data library and existing relevant database respectively and executes the experiment knot of data mining process The schematic diagram that fruit is arranged.

Here, in the case where existing relevant database, it is different from ontology data, by minimum support (minimum Support) value is only applicable in 0.6, so that experiment is possible, if becoming higher, can not be tested, so that experiment is only held Row is until the place.

Specifically, referring to Figure 41, Figure 41 is by graph representation with the ontology data of the embodiment of the present invention and existing Relevant database critical (the minimum support threshold) value of minimum support the memory that is changed of variation The schematic diagram of the change procedure of usage amount.

In other words, the quantity of affairs is limited to 100 and to the variation institute with minimum support value by the present inventor The memory usage amount of change, frequent mode quantity execute the time etc. and are investigated, firstly, according to the most ramuscule between two data How the variation observation memory usage amount for holding angle value changes, the results showed that, as shown in figure 41, the memory usage amount between two data It is similar.

Therefore, by experimental result as shown in figure 41 it is found that in the case where affairs are few, compared with existing relationship type number According to the data in library, the memory usage amount of the data of ontology database according to an embodiment of the present invention is slightly a little less, but compares class Seemingly.

It then, is that the quantity of affairs is also limited to 100 and by graph representation with the present invention referring to Figure 42, Figure 42 The number for the frequent mode that the variation of the minimum support value of the ontology data of embodiment and existing relevant database is changed Measure the schematic diagram of variation.

Here, the experiment is basis using the data of which kind of structure as experimental data, and in ontology database and generally Frequent mode structure is with progress lower premised on different form between database, as shown in figure 42, if observation ontology data The pattern count amount change procedure that takes place frequently between library and general database is then confirmed using ontology data, compared with General database take place frequently structure occur it is more.

The result is in the case where ontology database, because having the level knot of more structuring compared with general database Structure, so can produce the structure that more takes place frequently.

In other words, during component ontology database as described above, because being made by ontology inference process The relationship of each class mutually carries out structure connection, so can have the structure that more takes place frequently compared with general data, accordingly, according to this Invention, because can provide the structure that more takes place frequently, can be mentioned in the case where providing search result by searching for frequent mode For better search result.

Then, referring to the ontology data and existing pass that Figure 43, Figure 43 are by graph representation with the embodiment of the present invention It is the schematic diagram of the variation for the execution time (run time) that the variation of the minimum support of type database is changed.

In the experiment, transactions are also equally limited to 100 and are tested, as shown in figure 43, compared with existing Relational data, the execution time of ontology database is shorter.

Herein, it should be noted that confirm, in the case where relational data, when supporting angle value small and most Execution time when big sharply increases, but in contrast to this, in the case where ontology database, execute the variation of time Journey needs the similar execution time with the variation of minimum support value together.

Then, referring to Figure 44, Figure 44 is to be indicated with chart mode with the ontology data of the embodiment of the present invention and existing The schematic diagram of the variation for the execution time that the variation of the affairs of relevant database is changed.

In other words, as shown in figure 44, minimum support is fixed on 0.4, and while changing the quantity of affairs, table Show the variation of the execution time of ontology data according to an embodiment of the present invention and existing relevant database, the results showed that with Difference occurs for the increase of transactions, the execution time between two data, and compared with the data of existing relevant database, this The execution time of the data of the ontology database of inventive embodiments is shorter.

Therefore, compared with using existing relevant database the case where, utilizes the ontology database of the embodiment of the present invention Information retrieval, it is shorter the time required to retrieval in the retrieval request for receiving user.

Then, referring to Figure 45, Figure 45 be the ontology database that is compared with chart mode indicate with the embodiment of the present invention with The schematic diagram of the variation for the memory usage amount that the variation of the affairs of existing relevant database is changed.

Here, as shown in figure 45 the result is that minimum support is fixed on 0.4 and changes transactions while progress Experiment as a result, and it is as shown in figure 45, change changed memory usage amount with the quantity of affairs between two data, compared with The existing relevant database of utilization data the case where, using the ontology database of the embodiment of the present invention data the case where Middle memory usage amount is less.

In addition, compared with transactions to be limited to 100 situations, the situation few compared with the quantity of affairs, in affairs Quantity it is more when, compared with using existing relevant database data the case where, utilization is according to an embodiment of the present invention Memory usage amount in the case where the data of ontology database is more effective.

Accordingly, it is considered to the maximum problem for the information retrieval that can be used as in mobile environment mobile device capacity not Foot, it is according to an embodiment of the present invention using ontology database the case where it is more effective.

Then, illustrate the information retrieval to ontology database according to an embodiment of the present invention and existing relevant database The result tested of accuracy.

Firstly, Figure 46 is the variation table indicated using chart mode with the affairs as user query mode referring to Figure 46 Show the schematic diagram of the variation of the accuracy of recall precision.

In other words, the present inventor is using as the semantic data mining algorithm for being suitable for described the embodiment of the present invention FP-growth algorithm, measurement indicate the accuracy (Accuracy) of the accuracy of the information retrieval to experimental data.

Measurement result shows as shown in figure 46, ontology database according to an embodiment of the present invention and existing relationship type number Increase according to the accuracy in library with the increase of transactions.

But if searching system to existing portal or website relevant to going sightseeing and pass through the process The accuracy of the mobile message searching system for being applicable in semantic data method for digging of embodiment building according to the invention is compared Compared with can then find that the accuracy of existing system is obviously relatively low.

This is because existing system is provided and is wished with user when user inputs required search key to retrieval window The unrelated search result of the retrieval content of prestige.

Then, in order to the mobile letter for being applicable in semantic data method for digging according to the embodiment of the present invention of the explanation The difference of breath searching system and existing searching system is compared the accuracy of rear observed information retrieval, illustrates to according to this Input knot when going sightseeing ground in the mobile message searching system of inventive embodiments and existing portal to retrieval window respectively The result that fruit picture is compared.

It in other words, is to be shown when " chicken Longshan " is input to retrieval window by existing portal referring to Figure 47, Figure 47 Show the schematic diagram of the picture of search result.

Specifically, as shown in figure 47, the search result of existing system is, when the sightseeing to retrieval window input chicken Longshan When ground, although include in the content shown as search result value to practical relevant content of going sightseeing, it is completely irrelevant Content is also used as search result to show, at the same time, compared with computer picture, provides in much smaller smart phone picture excessive Search result, therefore, user always rolling picture while search result is confirmed, it is very inconvenient.

Therefore, it to solve the defect, needs only accurately advantageously to provide retrieval desired by user in search result As a result searching system, for this purpose, in the present invention, to solve the problems, such as the existing searching system, proposing and realizing to utilize language The mobile message searching system of adopted data digging method.

Specifically, Figure 48 is to retrieve in mobile message searching system according to an embodiment of the present invention referring to Figure 48 The schematic diagram of the picture of search result is shown when window input " chicken Longshan ".

In other words, in existing searching system, the content unrelated with chicken Longshan is provided to result screen, thus in the presence of The defect of information needed need to be confirmed one by one and selected to check by user, but as shown in figure 48, in shifting according to an embodiment of the present invention In dynamic information retrieval system, the identical chicken of the value of search key is inputted with to the retrieval window of existing searching system in input When Longshan, only by accurate value, value is provided as a result.

In addition, Figure 49 is defeated in mobile message searching system according to an embodiment of the present invention to be illustrated respectively in referring to Figure 49 When entering search key, by explanation relevant to search result and the picture for providing surrounding area map together as end value Schematic diagram.

As shown in figure 49, if can not only be provided using mobile message searching system according to an embodiment of the present invention Search result to the search key inputted, moreover, being for example worth offer together as a result includes to related ground of going sightseeing The map of surrounding area including the relevant informations such as more detailed description and periphery restaurant, therefore, compared with existing searching system The accuracy of its search result is higher.

Therefore, it can be realized in mobile environment using context aware ontology and be excavated based on semantic data by the process The mobile message searching system of method.

In other words, according to the present invention, it is possible to provide a kind of context aware ontology implementation method, the context aware ontology are real Existing method pays close attention to information service for providing user, wherein firstly, in order to store context data, by including that portal exists It introduces to the sightseeing in interior each area website and collects sightseeing information, collected information is pre-processed and is stored in data In library, and the relationship on each sightseeing ground of setting and the ontology inference mistake of attribute etc. are executed using the sightseeing database stored in this way Journey, to construct sightseeing ontology database.

In addition, according to the present invention, it is possible to provide a kind of user's concern information retrieval service offer excavated using semantic data Method, wherein in order to more rapidly accurately provide information to user using the sightseeing ontology database constructed by the process Search result is applicable in the data digging method based on existing FP-growth algorithm, thus by comparing user from the user Query pattern and the query pattern being stored in sightseeing ontology database, the inquiry to structure similar or identical can provide rapidly Search result.

At the same time, according to the present invention, it is possible to provide a kind of mobile message retrieval service is utilizing the context aware sheet While body and semantic data excavate the more succinct accurate search result for providing and being suitable for mobile environment, can execute together to The Map Services of the position of information retrieval result needed for the current location at family or user.

In addition, according to the present invention, in ontology inference process, being arranged by attribute setting and relationship so that synonym or class Processing like word, consonant word is possible, to can input various search keys during being retrieved, and can be by suitable It is solved that intelligence can not be overcome because providing unnecessary search result in existing retrieval service with semantic data digging method The problem of picture of terminating machine small defect.

So far, the context aware sheet according to an embodiment of the present invention based on context aware can be realized by the process Body and the user's concern information service for utilizing it.

Here, described the embodiment of the present invention in case where providing and have the concern information service constituted as follows is right It is described with: in order to predict and provide user to the concern information on ground of going sightseeing, constructing the situation sense about ground of going sightseeing Know ontology (Context Awareness Ontology), and semantic data mining process is executed based on the context aware ontology (Semant ic Data Mining) is provided a user more acurrate succinct as a result, to the offer more suitable for mobile environment It is illustrated for concern information service, but the present invention is not limited to this, in other words, the present invention is not only suitable for institute as above The concern information to ground of going sightseeing stated, moreover, the service to the concern information in other any regions can be suitable for as needed.

Therefore, it is realized by the process according to the present invention for providing user's concern information based on context aware Context aware ontology implementation method.

In addition, as described above, being realized by the process according to the present invention for providing the user based on context aware The context aware ontology implementation method for paying close attention to information provides a kind of user pass for providing based on context aware according to the present invention Infuse the context aware ontology implementation method of information service, wherein of interest based on context aware technological prediction user by building Content, and more acurrate prediction user is in the situation ontologies model for where which type of needing retrieve, and according to the feelings Border ontology model predicts the concern information of user, user's information of concern is only provided for the retrieval request of user, so as to mention For more acurrate brief search result, thus solve to provide a large amount of unnecessary search result for the retrieval request of user, To force what user filtered information needed again to be present in the problems in existing search method.

At the same time, according to the present invention, it provides a kind of for providing the situation for paying close attention to information of the user based on context aware Ontology implementation method is perceived, semantic data mining algorithm (the Semantic Data Mining based on situation ontologies is utilized Algorithm more acurrate brief search result is provided) so as to overcome the limited picture of the mobile device as smart phone The defect in face and few memory.

Especially, according to the present invention, it provides a kind of for providing the situation sense for paying close attention to information of the user based on context aware Ontology implementation method is known, while providing the search result for the retrieval request information of user, using being built in movement The GPS and map function of equipment provide the current location of user and the position of retrieval information needed for user with search result together It sets, to only can not only be easy to find the content of information needed by primary retrieval, additionally it is possible to be easy to find position.

By embodiment present invention as described above to according to the present invention for providing the user based on context aware The detailed content for paying close attention to the context aware ontology implementation method of information is illustrated, but the present invention is not limited to be recorded in institute State the content of embodiment, through the invention belonging to technical field the technical staff with general knowledge, can it is as needed and It is natural that other a variety of causes, which can carry out various amendments, change, combination and replacement etc.,.

Claims (9)

1. a kind of context aware ontology implementation method also provides not to solve to be directed to the retrieval request of user to the user Necessary search result, to force user to filter again and select information needed is present in asking in existing searching system Topic is only provided the concern information of the user as search result based on context aware technology, thus compared with existing retrieval System can provide more acurrate brief search result, and the context aware ontology implementation method includes:

User pays close attention to information database construction step, the information of the region-of-interest for the user is collected, to construct number According to library；

Ontology inference step, setting are stored in relationship and category between each data in user's concern information database Property, the user pays close attention to information database and constructs in user concern information database construction step；And ontology data Library construction step, according to the relationship for each data being arranged in the ontology inference step and attribute building ontology data Library；

In the ontology by the contextual information of the user be expressed as include user identifier U, spatial information L, temporal information T and The space contextual information O={ U, L, T, S } when having four-dimensional of information on services S relevant to the user；

The temporal information is the information divided by preset certain time spacing, and the spatial information is that belonging Europe is several The general location information of the coordinate (xi, yi) of Reed spatially, and the coordinate respectively indicates the user and is presently in Area longitude and latitude value；

In the database sharing step, as the concern information of the user, the letter to ground of going sightseeing is collected by each area Breath, as most higher level's class designated root class, is appointed as subclass for each area, by the sightseeing to each area title is specified For the subclass to each area, by the facility in each sightseeing ground or the affiliated detailed description for going sightseeing ground is appointed as to institute The subclass for stating ground title of going sightseeing, to each area or the synonym being arranged with going sightseeing for synonym or consonant retrieval Class relationship, to construct the sightseeing database for constituting hierarchical structure.

2. context aware ontology implementation method according to claim 1, it is characterised in that:

The sightseeing database includes id, add_1, add_2, add_3, add_4, add_5, tour_site, syn, lat, long Multiple domains, the domain add_1, the domain add_2, the domain add_3, the domain add_4 and the add_5 domain representation are each The address on sightseeing ground, the domain tour_site are stored with sightseeing ground title, and the domain syn is for synonym or consonant word The domain of processing, the domain lat and the domain long respectively indicate latitude and longitude required when setting map.

3. context aware ontology implementation method according to claim 2, it is characterised in that:

In the ontology inference step, determined according to " Place ", " Attraction ", " Resource ", " Activity " each The attribute of a data, " Place " are made of the title in the area to be gone sightseeing, and " Attraction " is by each area Interior sight-seeing resort title is constituted, and " Resource " is by the building or remains, traces, seashore and meal in each sightseeing ground Shop is constituted, and " Activity " includes " the Cultural activity " for indicating cultural activity, the behavior for indicating ornamental landscape " Scenery view ", indicate " the Dining out " of behavior that dines out, to indicate the user in each sightseeing Practical the implemented action on ground.

4. context aware ontology implementation method according to claim 3, it is characterised in that:

In the ontology inference step, the relation on attributes of by described in " Place " and " Attraction " is appointed as " hasAttraction " and " IsAttractionOf ", the attribute of " Attraction " and " Resource " by described in close It means and is set to " hasResource " and " IsResourceOf ", and the category of " Activity " and " Resource " by described in Sexual intercourse is appointed as " hasActivity " and " IsActivityOf ".

5. context aware ontology implementation method according to claim 4, it is characterised in that:

It is considered as there are when multiple relations on attributes in the data for constituting each class with multiple category in the ontology inference step Sexual intercourse.

6. context aware ontology implementation method according to claim 1, it is characterised in that:

In the ontology inference step, executed using Protege_4.0.2version.

7. context aware ontology implementation method according to claim 1, it is characterised in that:

In the ontology database construction step, the processing for constructing the ontology database is executed using OWL, RDF and RDFS.

8. a kind of user based on context aware pays close attention to information service providing method, it is characterised in that:

Utilize the context aware that context aware ontology implementation method constructs documented by any one by claim 1 to 7 Ontology only provides the concern information of user as search result, to provide compared with existing search method more acurrate brief Search result, therefore, more suitable for the mobile environment including smart phone or tablet computer.

9. a kind of user based on context aware pays close attention to information service system, it is characterised in that:

Utilize the context aware that context aware ontology implementation method constructs documented by any one by claim 1 to 7 Ontology only provides the concern information of user as search result, to provide compared with existing searching system more acurrate brief Search result, therefore, more suitable for the mobile environment including smart phone or tablet computer.