Matches in ScholarlyData for { ?s ?p ?o. }
- item-4 type ListItem.
- item-4 hasContent dipanjan-chakraborty.
- item-4 next item-5.
- item-4 previous item-3.
- item-5 type ListItem.
- item-5 hasContent amit-a-nanavati.
- item-5 previous item-4.
- 224 creator le-minh-nguyen.
- 224 creator susumu-horiguchi.
- 224 creator xuan-hieu-phan.
- 224 type InProceedings.
- 224 label "Learning to Classify Short and Sparse Text & Web with Hidden Topics from Large-scale Data Collections".
- 224 sameAs 224.
- 224 abstract "This paper presents a general framework for building classifiers that deal with short and sparse text & Web segments by making the most of hidden topics discovered from large-scale data collections. The main motivation of this work is that many classification tasks working with short segments of text & Web, such as search snippets, forum & chat messages, blog & news feeds, product reviews, and book & movie summaries, fail to achieve high accuracy due to the data sparseness. We, therefore, come up with an idea of gaining external knowledge to make the data more related as well as expand the coverage of classifiers to handle future data better. The underlying idea of the framework is that for each classification task, we collect a large-scale external data collection called "universal dataset", and then build a classifier on both a (small) set of labeled training data and a rich set of hidden topics discovered from that data collection. The framework is general enough to be applied to different data domains and genres ranging from Web search results to medical text. We did a careful evaluation on several hundred megabytes of Wikipedia (30M words) and MEDLINE (18M words) with two tasks: "Web search domain disambiguation" and "disease categorization for medical text", and achieved significant quality enhancement.".
- 224 hasAuthorList authorList.
- 224 hasTopic World_Wide_Web.
- 224 isPartOf proceedings.
- 224 keyword "Sparse text & Web classification".
- 224 keyword "Web data analysis".
- 224 keyword "hidden topic analysis".
- 224 keyword "medical text analysis & classification".
- 224 title "Learning to Classify Short and Sparse Text & Web with Hidden Topics from Large-scale Data Collections".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasItem item-2.
- authorList hasItem item-3.
- authorList hasLastItem item-3.
- item-1 type ListItem.
- item-1 hasContent xuan-hieu-phan.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent susumu-horiguchi.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent le-minh-nguyen.
- item-3 previous item-2.
- 226 creator jianfeng-du.
- 226 creator yi-dong-shen.
- 226 type InProceedings.
- 226 label "Computing Minimum Cost Diagnoses to Repair Populated DL-based Ontologies".
- 226 sameAs 226.
- 226 abstract "Ontology population is prone to cause inconsistency because the populating process is imprecise and the populated data may conflict with the original data. By assuming that the intensional part of the ontology is fixed and each removable ABox assertion is given a removal cost, we intend to repair an inconsistent DL-based ontology by deleting a subset of removable ABox assertions in which the sum of removal costs is minimum. We call such subset a minimum cost diagnosis. We show that computing a minimum cost diagnosis for a DL-Lite ontology cannot be accomplished in polynomial time w.r.t.\ data complexity unless P=NP. Nevertheless, we present a feasible computational method for more general (i.e., SHIQ) ontologies. It transforms a SHIQ ontology to a set of propositional programs in polynomial time w.r.t.\ data complexity, thus reducing the original problem into a set of subproblems that are independent of each other, where each subproblem computes a certain model and is solvable in logarithmic calls to a standard SAT solver. Experimental results show that the proposed method can practically handle moderately complex ontologies with thousands of removable ABox assertions.".
- 226 hasAuthorList authorList.
- 226 hasTopic World_Wide_Web.
- 226 isPartOf proceedings.
- 226 keyword "Description Logics".
- 226 keyword "Diagnosis".
- 226 keyword "Disjunctive Datalog".
- 226 keyword "Ontology Population".
- 226 keyword "Partitioning".
- 226 title "Computing Minimum Cost Diagnoses to Repair Populated DL-based Ontologies".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasItem item-2.
- authorList hasLastItem item-2.
- item-1 type ListItem.
- item-1 hasContent jianfeng-du.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent yi-dong-shen.
- item-2 previous item-1.
- 233 creator atsushi-fujii.
- 233 type InProceedings.
- 233 label "Modeling Anchor Text and Classifying Queries to Enhance Web Document Retrieval".
- 233 sameAs 233.
- 233 abstract "Several types of queries are widely used on the World Wide Web and the expected retrieval method can vary depending on the query type. We propose a method for classifying queries into informational and navigational types. Because terms in navigational queries often appear in anchor text for links to other pages, we analyze the distribution of query terms in anchor texts on the Web for query classification purposes. While content-based retrieval is effective for informational queries, anchor-based retrieval is effective for navigational queries. Our retrieval system combines the results obtained with the content-based and anchor-based retrieval methods, in which the weight for each retrieval result is determined automatically depending on the result of the query classification. We also propose a method for improving anchor-based retrieval. Our retrieval method, which computes the probability that a document is retrieved in response to the given query, identifies synonyms of query terms in the anchor texts on the Web and uses these synonyms for smoothing purposes in the probability estimation. We use the NTCIR test collections and show the effectiveness of individual methods and the entire Web retrieval system experimentally.".
- 233 hasAuthorList authorList.
- 233 hasTopic World_Wide_Web.
- 233 isPartOf proceedings.
- 233 keyword "Web retrieval".
- 233 keyword "anchor text".
- 233 keyword "query classification".
- 233 title "Modeling Anchor Text and Classifying Queries to Enhance Web Document Retrieval".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasLastItem item-1.
- item-1 type ListItem.
- item-1 hasContent atsushi-fujii.
- 247 creator christian-stahl.
- 247 creator dieter-koenig.
- 247 creator karsten-wolf.
- 247 creator niels-lohmann.
- 247 creator simon-moser.
- 247 type InProceedings.
- 247 label "Extending the Compatibility Notion for Abstract WS-BPEL Processes".
- 247 sameAs 247.
- 247 abstract "WS-BPEL defines a standard for executable business processes. Executable processes are business processes which can be automated through an IT infrastructure. The WSBPEL specifcation also introduces the concept of abstract processes: In contrast to their executable siblings, abstract processes are not executable and can have parts where business logic is disguised. Nevertheless, the WS-BPEL specification introduces a notion of compatibility between such an under-specified abstract process and a fully specified executable one. Basically, this compatibility notion defines a set of syntactical rules that can be augmented or restricted by profiles. So far, there exists two of such profiles: the Abstract Process Profile for Observable behavior and the Abstract Process Profile for Templates. None of these profiles defines a concept of behavioral equivalence. Therefore, both profiles are too strict with respect to the rules they impose when deciding whether an executable process is compatible to an abstract one. In this paper, we propose a novel profile that extends the existing Abstract Process Profile for Observable Behavior by defining a behavioral relationship. We also show that our novel profile allows for more flexibility when deciding whether an executable and an abstract process are compatible.".
- 247 hasAuthorList authorList.
- 247 hasTopic World_Wide_Web.
- 247 isPartOf proceedings.
- 247 keyword "Abstract Profile".
- 247 keyword "Compliance".
- 247 keyword "Petri nets".
- 247 keyword "WS-BPEL".
- 247 title "Extending the Compatibility Notion for Abstract WS-BPEL Processes".
- authorList type List.