In as long as mation Discovery on Vertical Domains Need as long as In as long as mation Discovery Strengths in addition to Limitations of Current Approaches Research Objective Specific Domains Studied (or being studied)

In as long as mation Discovery on Vertical Domains Need as long as In as long as mation Discovery Strengths in addition to Limitations of Current Approaches Research Objective Specific Domains Studied (or being studied) www.phwiki.com

In as long as mation Discovery on Vertical Domains Need as long as In as long as mation Discovery Strengths in addition to Limitations of Current Approaches Research Objective Specific Domains Studied (or being studied)

Aikman, Troy, Host has reference to this Academic Journal, PHwiki organized this Journal In as long as mation Discovery on Vertical Domains Vagelis HristidisAssistant ProfessorSchool of Computing in addition to In as long as mation SciencesFlorida International University (FIU), MiamiNeed as long as In as long as mation DiscoveryAmount of available data increasesNeedle in the haystack problemSome applications:WebDesktop searchData WarehousingBibliographic databaseHomes, cars search, e.g., realtor.com, autotrader.comScientific domains, e.g., genes, proteins, publications in biology, elements in addition to interactions of components in chemistryPatient hospitalizations, physician info, procedure outcomes in hospitalsVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 2Strengths in addition to Limitations of Current Approaches Vagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 3Web Search+ Scalability+ H in addition to le free text+ Exploit content in addition to link structure to achieve ranking+ Simple keyword queries- Limited query expressive power- Generic, domain-independent ranking algorithms- Return pages, not answersDatabase Querying+ Efficient+ H in addition to le structured data+ Well-defined theory in addition to answers- Must learn query language, e.g. SQL- No automatic ranking of resultsKeyword Search in Databases + Simple keyword queries + exploit links (e.g., primary- as long as eign keys) – Generic ranking – typically size of result – No domain semantics

Boston Architectural Center US www.phwiki.com

This Particular University is Related to this Particular Journal

Research ObjectiveAllow effective in addition to efficient in as long as mation discovery on vertical domainsStrategy:Exploit associations between entitiesModel domain semantics, e.g., patient entity is critical as long as medical practitioner, but not as long as biologistModel users of a domainUse knowledge of domain experts, in addition to existing knowledge structures (e.g., domain ontologies)Exploit user feedbackGo beyond plain keyword search. Explore best search interface as long as each domain, e.g., faceted searchVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 4Specific Domains Studied (or being studied)Products marketplaceBiological databasesClinical databasesBibliographicPatentsVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 5Specific Domains Studied (or being studied)Products marketplaceBiological databasesClinical databasesBibliographicPatentsVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 6

Products MarketplaceProject started while visiting Microsoft Research at Redmond, in Summer 2003SQL Returns Unordered Sets of ResultsOverwhelms Users of In as long as mation Discovery ApplicationsHow Can Ranking be Introduced, Given that ALL Results Satisfy QueryVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 78Products Marketplace (cont’d) Example – Realtor DatabaseHouse Attributes: Price, City, Bedrooms, Bathrooms, SchoolDistrict, Waterfront, BoatDock, YearQuery: City =`Seattle’ AND Waterfront = TRUEToo Many Results!Intuitively, Houses with lower Price, more Bedrooms, or BoatDock are generally preferableVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 9Products Marketplace (cont’d) Rank According to Unspecified Attributes [VLDB’04,TODS’06]Score of a Result Tuple t depends onGlobal Score: Global Importance of Unspecified Attribute ValuesE.g., Newer Houses are generally preferredConditional Score: Correlations between Specified in addition to Unspecified Attribute ValuesE.g., Waterfront BoatDock Many Bedrooms Good School DistrictVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains

10Products Marketplace (cont’d) Key ProblemsGiven a Query Q, How to Combine the Global in addition to Conditional Scores into a Ranking Function. Use Probabilistic In as long as mation Retrieval (PIR).How to Calculate the Global in addition to Conditional Scores. Use Query Workload in addition to Data.Vagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains Products Marketplace (cont’d) Other ProjectsSelect the best attributes to output – attribute ordering problem [SIGMOD’06]E.g., Color is important as long as sports cars but not much as long as family carsProduct Advertising: Select best attributes to display as long as a product to maximize its visibility among its competitors [ICDE’08, TKDE’09]Use past query workloadMaximize number of past queries as long as which the product is returnedVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 11Specific Domains Studied (or being studied)Products marketplaceBiological databasesClinical databasesBibliographicPatentsVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 12

Biological Databases [EDBT’09]With University of Maryl in addition to Intuitive but powerful query language, based on soft (ranking) in addition to hard (pruning) filtersGoal is to improve the user experience of users of PubMedExploit associations between entities (genes, proteins, publications)Example of Query: Find the most important publications on “cancer” that are related to the “TNF” gene through a protein.Vagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 13Results Navigation in PubMed with BioNav [ICDE’09, TKDE’10]With SUNY Buffalo.Most publications in PubMed annotated with Medical Subject Headings (MeSH) terms.Present results in MeSH tree.Propose navigation model in addition to smart expansion techniques that may skip tree levels.Vagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 14BioNav: Exploring PubMed Results15Vagelis Hristidis, Searching in addition to Exploring Biomedical Data

BioNav: Exploring PubMed ResultsReveal to the user a selected set of descendent concepts that:Collectively contain all resultsMinimize the expected user navigation costNot all children of the root are necessarily revealed as in static navigation.16Vagelis Hristidis, Searching in addition to Exploring Biomedical DataBioNav Evaluation17Vagelis Hristidis, Searching in addition to Exploring Biomedical DataSpecific Domains Studied (or being studied)Products marketplaceBiological databasesClinical databasesBibliographicPatentsVagelis Hristidis – FIU – In as long as mation Discovery on Vertical Domains 18

XOntoRank: Use Ontologies to Search Electronic Medical Records [ICDE’09]With Miami Children’s Hospital, Indiana University School of Medicine, IBM Almaden.Latest EMR as long as mat: HL7 CDA – XML-basedAlgorithm to enhance keyword search using ontological knowledge (e.g., SNOMED)19Vagelis Hristidis, Searching in addition to Exploring Biomedical Data20SAMPLE CDA FRAGMENTVagelis Hristidis, Searching in addition to Exploring Biomedical DataXOntoRank: Example 1q = {“bronchitis”, “albuterol”}21Vagelis Hristidis, Searching in addition to Exploring Biomedical Data

XOntoRank: Example 2q = {“asthma”, “albuterol”}result = 22Vagelis Hristidis, Searching in addition to Exploring Biomedical DataXOntoRankA CDA node may be associated to a query keyword w through ontology.XOntoRank first assigns scores to ontological conceptsOntoScore OS(): Semantic relevance of a concept c in the ontology to a query keyword w.Then, given these scores, assign Node Scores NS() to document nodesOther aggregation functions are possible.23Vagelis Hristidis, Searching in addition to Exploring Biomedical DataComputing OntoScore of Concept Given Query KeywordThree ways to view the ontology graph:As an unlabeled, undirected graph.As a taxonomy.As a complete set of relationships.24Vagelis Hristidis, Searching in addition to Exploring Biomedical Data

Aikman, Troy Troy Aikman Show, The Host www.phwiki.com

Authority Flow Ranking in EMRsA subset of the electronic health record dataset.Work under submission.

Aikman, Troy Host

Aikman, Troy is from United States and they belong to Troy Aikman Show, The and they are from  Scottsdale, United States got related to this Particular Journal. and Aikman, Troy deal with the subjects like Football

Journal Ratings by Boston Architectural Center

This Particular Journal got reviewed and rated by Boston Architectural Center and short form of this particular Institution is US and gave this Journal an Excellent Rating.