【正文】 1Concept 2Attributes 2Operations 2Concept 3Attributes 3Operations 3Concept nAttributes nOperations nConcept 4Attributes 4Operations 4Concept 5Attributes 5Operations 5Concept 6Attributes 6Operations 6………Concepts SetAssociation GeneralizationAggregation15The fourlayered ObjectOriented OntologyDomainCategory 1 Category 2 Category kConcept 3Attributes 3Operations 3Concept 1Attributes 1Operations 1Concept 2Attributes 2Operations 2Concept nAttributes nOperations nConcept 4Attributes 4Operations 4 ClasslayerInstance 3Attributes 3Operations 3Instance 1Attributes 1Operations 1Instance 2Attributes 2Operations 2Instance mAttributes mOperations mInstance 4Attributes 4Operations 4 InstancelayerAssociationGeneralizationAggregationInstanceof16The fourlayered News Ontology (cont.)DomainEvent 1 Event kEvent 3Event 2 …Candidate Chinese Terms… … ………… Extended ConceptRelation…Category 1 Category 2 Category qAssociationOperationTemplatesAttributesEvent Concept 1OperationAttributesObject Concept 1, mOperationAttributesEvent Concept 1, 1OperationAttributesEvent Concept POperationAttributesEvent Concept P,1OperationAttributesObject Concept P,n 17The fourlayered News Ontology發(fā)佈、表示導(dǎo)致、造成、帶來氣象影響向、往遠(yuǎn)離、移動帶來、引進(jìn)氣象報導(dǎo) 氣象百科 天文. . . . . . 寒流 颱風(fēng) 降雨. . . . . . . . . . . . RelationAssociation表示、警告、評估型態(tài)：預(yù)報人員、 天氣圖中央氣象局 /氣象局來襲、形成、登陸編號 :***(Neu)號中心位置：:***(Nc)(Ncd)(Neu)(Nf)強(qiáng)度 :輕度颱風(fēng)型態(tài) :暴風(fēng)圈 颱風(fēng)發(fā)生、襲擊、增加降雨量 ***(Neu)公釐累積雨量 ***(Neu)公釐種類 :大雨、陣雨、 大雷雨、豪雨 、豪大雨型態(tài)：雨量、打雷降雨移動、靠近、前進(jìn)方向 :東方、南方 西北方、東 南方移動方向接近、影響、流動型態(tài) :西南氣流、 冷氣流氣流避風(fēng)、休耕型態(tài) :漁港、農(nóng)田 、農(nóng)作物、 魚貨量農(nóng)林漁牧業(yè)呈現(xiàn)、滯留、徘徊區(qū)域 :山區(qū)、平地、 臺灣、中部、 東半部各縣市：臺北市、臺 南縣海域 :東海、南海海岸 :西海岸、沙岸地區(qū)來襲、形成、登陸型態(tài) :水災(zāi)、旱象、 土石流、山崩 、洪水、房屋 倒塌、河水暴 漲、落石、雷 擊、霜害 災(zāi)害增強(qiáng)為、逼近型態(tài) :副熱帶高氣 壓、熱帶性 低氣壓氣壓發(fā)生導(dǎo)致造成注意、受困型態(tài) :人數(shù)民眾 /人民提醒根據(jù)、開始型態(tài) :最近、昨日 今日、白天 午後時間影響恢復(fù)出現(xiàn)、發(fā)生18Fuzzy Ontology (cont.)DomainCategory 2 ……………C : ConceptA : AttributeO : OperationCategory 1 Category 3 Category kClasslayer{C1。μC1E1,μC1E2,…,μ C1Ep}AC11,AC12 ,…,A C1q1{Cm。μC2E1,μC2E2,…, μC2Ep}AC21,AC22 ,…,A C2q2{C3。μC4E1,μC4E2,…, μC4Ep}AC41,AC42,…,A C4q4OC41,OC41,…,O C4q4{C5。WeightingPresentPresent BrowsingBehaviorConcepts and Weightsof BrowsedContentSequence of User Browsing Behaviors1 s2 s3 sn…User Behavior amp。 Browsing Content Analysis…… … … …Additional Weighting of Related Sequence ConceptsDomainOntologyPersonalOntologyPersonal Ontology31User Behavior Analysiso In order to find out user’s favor tendency, the first job is analyzing the habitual behavior of reading.o Consider two features: reading time and reading frequency. o Consider reading time is related with content length, change the feature to32BrowsingFrequencyBrowsingTimeContentLengthPersonal Log FilesRepositoryBrowsedContentsRepositoryFeatureDataData Sorting…Data ClusteringSortedDataFunctionalizationFunctionof FeatureDataFeature Data FunctionalizationFeature DataProcessingFrequencyFeature DataBrowsing TimeFeature DataFeature Data FunctionalizationFunction ofBrowsingFrequencyFunction ofBrowsingTimePersonal Ontology33Question amp。 othero YesNo question…etc.o Answer analysisn Question typeo 5W1Hn Domaino Domain knowledge34Question amp。AKMworkflowOntology Qamp。 learning processQuestion Answering SubsystemKnowledge Extraction SubsystemReceive User queryDocumentsReturn AnswerOntology supervision Question amp。 Answer Knowledge Base (cont.)o Domain ontologyn Objectoriented ontology o Question ontologyn The knowledge of question domainn To Classify and extract questiono Answer ontologyn The knowledge map of Qamp。 Answer Knowledge Baseo Alternation Rulen Morphological n Lexical n Semantic o Ontology supervisionn Ontology managementn Ontology inference37IntereNews RetrievalAgent Fuzzy InferenceAgentChinese eNews SummaryChinese eNewsOntology…ChineseeNews SummaryRepositoryRealtime eNewsRepositoryeNews RepositoryGUIPOS Tagger(CKIP)Chinese Term FilterDocument Processing AgentOFEE AgentExtractedEventOntologyPDACell PhoneNotebookEvent Ontology FilterSent