• Save
Social Media Lab - Elsevier WebApp Concept Design Competition 2011
Upcoming SlideShare
Loading in...5
×
 

Social Media Lab - Elsevier WebApp Concept Design Competition 2011

on

  • 3,159 views

Dalhousie University

Dalhousie University
Social Media Lab
http://socialmedialab.ca/?page_id=3106

Statistics

Views

Total Views
3,159
Views on SlideShare
3,159
Embed Views
0

Actions

Likes
0
Downloads
0
Comments
0

0 Embeds 0

No embeds

Accessibility

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

Social Media Lab - Elsevier WebApp Concept Design Competition 2011 Social Media Lab - Elsevier WebApp Concept Design Competition 2011 Presentation Transcript

  • Team Member Names: Raheleh MAKKI, Ozge YELOGLU, Axel SOTO T M b N R h l h MAKKI, O g YELOGLU, A l Faculty of Computer Science, Dalhousie University, Halifax, Canada y p , y, , HOW?Abstract Corpus l Corpus-level Vi C level View l • Th Cube visualizes up to 200 documents, where similar The C b i li t d t h i il documents are located close to each other. Blob CubeOur application, BlobCube, provides two ways of visualizing pp , ,p y g • Similarity is defined based on the frequency of three keywords y q y ythe user’s search results: document-level and corpus-level user s selected by the user. uservisualization. Th d i li ti The document-level visualization, th Bl b tl l i li ti the Blob, • Keywords and full-text of documents are retrieved from theshows the degree of relevance of four keywords to each SciVerse Content API. S iV C t t APIdocument.document The corpus level visualization the Cube focuses corpus-level visualization, Cube,on the content similiarities among the retrieved documents. g • C l represents either t Color t ith type or subject area of the documents. bj t f th d t • Size represents either recency or relevance of the documents documents.ThisThi application t g t a user who wants a f t i ight of li ti targets h t faster insight fher search results. As a result, this application brings the WHY?scientific content and the ranked list representation together ranked-list • To get a faster understanding of how retrieved documentsby employing different visualization approaches. y p y g pp differ diff r among each other. ng h th r • To offer a more comprehensive representation of the search results than traditional ranked-list.Document-levelDocument level View HOW? • Each Blob summarizes the content of a document in h l b h f d Conclusions • Bl bC b BlobCube: terms of four keywords [1] [1]. • meets the users’ need of gaining valuable insight from a g g g • Keywords and full-text of documents are retrieved full text large set of documents in a shorter time time. from the SciVerse Content API API. • runs on Hub and Science Direct. • The full text of each article is used to calculate the term full-text • h a simple and flexible interface. has i pl d fl ibl i f frequencies of the f f q f h four k y keywords. d • Even though the Blob and Cube are not novel ideas our ideas, application employs these visualizations thorough SciVerse in pp p y g an innovative way way. TFi: Normalized term frequency of keyword i Requirements: q • Bl bC b requires a J BlobCube i Java Script visualization library, such S i t i li ti lib h as d3 [2]. [ ] WHY? • O ly Elsevier journal suscribers are able to use this Only El i j l ib bl hi application to its full potential potential. • To give the user a high-level summary of a document. g g y • Keyterm frequencies can only be calculated when the user y q y • T capture additional information not represented in To p ddi i li f i p di has access to full text of the documents full-text documents. the ranked-list (subtopics) ranked list (subtopics). [1] R.M. Rohrer, J.L. Sibert, D.S. Ebert.  A Shape based Visual Interface for  [1] R.M. Rohrer, J.L. Sibert, D.S. Ebert. “A Shape‐based Visual Interface for • The Blob shape makes identifying the salient keywords Text Retrieval , IEEE Computer Graphics and Applications, 19, 40 46, 1999. Text Retrieval”, IEEE Computer Graphics and Applications, 19, 40‐46, 1999. easy to the human eye [1] [1]. [2] d3:  Data driven documents . Free Java Script [2] d3: “Data‐driven documents”. Free Java Script library for visualization http://mbostock.github.com/d3/
  • Team Member Names: Laura Dukowski, Terry LeBlanc, Craig MacEachern School of Information Management, Library and Information Studies, Dalhousie University, Halifax, Canada Value for Users Application Description Target user group: Academic researchers collaborating on DataSnippets allows users to extract and share small portions of a single project. Snippets will be drawing data from user input and automatichyperlinked article content during the collaborative research •Keeps track of key article points; metadata extraction. User input will be limited to the choice ofprocess. These pieces of extracted text will hereafter be •Provides easy access to the source and context of the citation styles from the drop-down menu in the application, andreferred to as “snippets”. As the user highlights and extracts snippets; highlighting snippets from the article using the snippet-tool.article text, the snippets will appear on a clipboard feature, Export the snippets by clicking on the Google Documentsunder a citation listing for the corresponding article. Each •Saves time by eliminating the need to scroll through export button, which will be prompted by a pop-up window tosnippet will be hyperlinked to the Elsevier article in which it marked-up full text articles during research compilation; enter their username, password, and a filename.The developersappears. As the user extracts snippets from subsequent articles, m •Allows efficient sharing of user-identified key article of the application will need to be able to automatically harvestthese snippets will continue to appear on the clipboard. When points and article links. metadata from the articles. Developers could look tothe user wishes to end the research session, s/he may export applications like Cite Me! which already make use of the samethe contents of the clipboard to Google docs. The resulting file API calls to ScienceDirect that we would be using. The order ofuploaded to Google docs will contain the ordered list of Integration with article the metadata will depend on which formatting style is selected,snippets, retaining the hyperlinks to their corresponding but we would need the following metadata (if available):Elsevier articles. From Google docs, researchers will have the author(s), year published, title of article, journal name, volume,freedom to share key article points with colleagues and link issue, pages inclusive, digital object identifier (DOI).directly to the full text article quickly. Similar Applications Scivernote is the most similar application to Snippets, but lacks Snippets’ major features and exports to Evernote instead of Workflow Google Documents. Snippets allows extraction of highlighted text, and corresponding metadata. Scivernote exports parts of the article but not the passages that you want to use from the article. Snippet users may choose a formatting style prior to Click scissors Highlight desired text exporting their clipboard contents. Finally, Snippets allows icon to activate exporting to Google Documents; superior for sharing and editing than Evernote. The Cite Me! allows you to create a copy-and-paste bibliographic entry of the article you are viewing in either APA, AMA, MLA but does not copy Format styles highlighted text and hyperlink it to the corresponding article;Browse more articles Snippets move to nor does it export. Snippets has an integrated drop-down bar clipboard Article citation that allows you to choose any of the Cite Me! formatting, plus it allows you to save and hyperlink the highlighted passages, Snippets and export the entire list to Google Documents. Browsing complete Export to Google Docs APIs Snippet-mode To capture the metadata needed for export to Google Documents we will be using data pulled from the article using the SciVerse ScienceDirect API. To export the snippets we will be using the Export Google Documents List API.
  • MiSTeRy Visualizer Farhan Jamal Khan [farhan@cs.dal.ca] Mine Search Transform (MiSTeRy) Visualizer Ishan Patel [patel@cs.dal.ca] System Design Mine Search Semantic types Concepts related_to Is_a relations relations associated_with relations Concept 1 Concept 2 Query UMLS Semantic UMLS Concept 3 Entrez EUtilities Term 1 Network Metathesaurus Term 2 Results / Articles Set Results / Articles Set Term 3 Meta Map Concept 4 Semantic Techniques Term 4 Concept 1 Score [Semantic Type] Concept 2 Score [Semantic Type] Concept 3 Score [Semantic Type] PubMed Concept N Score [Semantic Type] Java Mine Search Transform Visualize Vaadin Transform Visualize Text Mining Database Searching Ranking Presentation MeSH Concepts Relevance Result 1 Result 2 Concept 1 Score Result 3 Title Concept 2 Publi ca Sema tion Info Concept 3 ntic Conc Type epts Result 4 Concept 4 Result 1 Result 2 List List Result 3 Clinical Text Query Query + Domain Domain Semantic Types/ PubMed Database Relevance Relevance Relevance Result 4 Sorting/ Concepts/ Graph Meta Map Scores MedLine Index Categorization Interactive Scoring Semantic MeSH Mine, Search, Transform (MiSTery) Visualizer Type Concepts Ranked Results Ranked Results Ranked Results Ranked Results Ranked Results Ranked Results Application Design MiSTeRy Visualizer Abstract: Search Results Page Graphical Representation Classical Keyword-based search has a number of limitations. These limitations include missing semantic information for terms, single-document retrieval, and lack of contextual ranking. Current research in tackling these constraints deals with concept-based information retrieval User Query Search and search. Mine, Search, Transform (MiSTeRy) Visualizer is a concept-based search application tion ssocia which utilizes the state-of-the-art research for incorporating semantic techniques in the Article Name Refine Results hasA domain of retrieving evidence-based literature from PubMed. The articles in PubMed are Score Authors, Publication Details Semantic Types Semantic Type, Concepts Type 1 indexed with MeSH (Medical Subject Headings) concepts. The Medical Subject Headings Type 2 Type 3 Semantic Type2 (MeSH®) thesaurus is a controlled vocabulary produced by the U.S. National Library of Medicine Article Name (NLM) and used for indexing, cataloging, and searching for biomedical and health-related Score Authors, Publication Details Semantic Concepts [370] Semantic Type3 Semantic Type, Concepts Concept 1 [500] information and documents. We have utilized the MeSH indexing of PubMed database to Sub Concept 1 Semantic Type1 improve the precision/recall ratio for the retrieved documents. The user of the system is a Sub Concept 2Ranked Articles Article Name Sub Concept 3 [100] researcher who has a set of terms and is looking for the literature relevant to those terms. The Score Authors, Publication Details Concept 2 Concept 1 Semantic Type, Concepts query provided by the user is expanded using the NLMs MetaMap tool. MetaMap uses UMLS Ranking Criteria Query Relevance Metathesaurus and Semantic Network to employ knowledge-intensive approach for identifying Article Name Concept 1 [150] nouns and corresponding scored concepts and semantic types from the provided text. The Score Domain Relevance Authors, Publication Details Query + Domain Relevance Semantic Type, Concepts Concept 1.1 [70] retrieved concepts are used to perform a query on PubMed to retrieve a set of articles. These Authors Concept 1 [50] Concept 1.2 [40] articles are ranked based on the commulative MeSH concept scores (query relevance) as well as Author 1 Article Name Concept 1.1 [30] on the basis of frequency of occurrence of a concept in retrieved set vs. the whole of PubMed Score Author 2 Authors, Publication Details Author 3 Semantic Type, Concepts Concept 1.2 [20] (domain relevance). Moreover, the concepts and semantic types of the MeSH terms in the Country Concept 2 [30] retrieved resultset are used to provide user with an option of refining (specializing/ Article Name Country 1 generalizing) the search and present a graphical summary of the results. Score Authors, Publication Details Country 2 Semantic Type, Concepts Country 3 Refine Concept 10 [120] Keywords: Keyword-based search, Concept-based search, MeSH, semantic search.
  • B N FT : E E IS B cue i ls rvd sntna e u acs t ifr t no a em o ea s Smpu po ie isa tn o s ceso nomai n tr r o smb l h t n r l se t nrsac igoh ro re t y o, e i omal p n o ee rhn te sucso t me y IE : DA u d rtn te em iei n td T eeoete i rq i d n esa d h tr s l ae .h rfr,h t e ur mi me eSinic tr l f n o tis eds e cag n hc cn k u d rtn ig h maeil t c t maei ot cnan l- c jro w i a ma e n esa dn te tr a i e a e p h a me cnu n a d rsrt g rcs,s eil fre d rw o ak x ein en h ed o smig n f t i poesep c l o ra es h lc e p r c i te l. u an ay e t ra a du d rtn a c nic o e d n n esa d si t e Smpu po ie isa t ceso nomain h t x lish s tr ,i ly g i ls rvd sntn acs t ifr t ta e pan tee emss i i o mp f n p p r y e ralrd cd a e ma b ge t e ue . y te rcs o ra ig n u d rtn ig c n c tr l h poes fe dn a d n esa dn si t maei . e a Ah ui iipo oe t a desre aa treet n o S p ot etr c ie, i te o l e r t s rp sd o d rs f p rmee slci fr u p rV co Mahn sw t h g as sc e o h o i rvn g n rlain efr n e n po iig rae isnivt t t iig e slcin fmpo ig e eai t p r ma c a d rvdn ge trne s ii o r nn steet . z o o t y a o T e n lclxrman h s o t zt n rbe ma e rde t ecn ag rh i rci l h ma yoae t e i tee pi ai po lms k ga in d se t loi msmpat a. mi o t c o H ui i(o ue Sin e e r t C mp tr c c) sc o mu e T e i p it fh po oe h ui iiten ls n f mo ecmpe i me sr t i rv h man on o te rp sd e r t sh ic i o a d lo lxt aue o mpo e sc uo n y g n rlain efr n eWe l ue i ltda n aig o mpo e aa tre rh fc n y e eai t p r ma c. as s s ae n e l t i rv p rmee sac efi i c z o e cmp rdt a e h ut e r sac ,n ic d a itni - eg tdcnr o mas fh mot o ae o n x a si gi e rh a d n l e n ne s yw ihe e t f s o te s v d u t e E pa ain T ue f n n o lt ifr t n o p rx x ln t : o s ot icmpeenomai t a po i o e o - o t m p itt rd c v lti . e a n t o tn adc sict n rbe fr o ai n pimu onso e u e oait We x mie w sa d r l s ai po lmso cmp r o , ly a fi o s a da py h h ui it bonomai a drt aeet p yilg c sict n [] n p lte e r t o iifr t s n ei llcr h s o y l s ai . sc c n o o a fi o 1 mae ouin, s gmeh d ce td rm e p r n e t slt sui o n to s rae fo x ei c. e F l x ln t n ik w we a l. m ul pa ai l : w . mpe o e o n x c HOW IIUS D: TS E Wh naue ra s peeo sinicmaeiln n s tr o smb l ed e n t n esa d te smperl hs e sre d a ic fce t traa d d a em ry o h o s o u d rtn ,h y i l ol i s cro o e ia dab xwia p a u d rh cro.h b xwicnana epa aino tetr o smb l usr v rt n o l p e r n e te usrT e o l o ti n x ln t fh em ry o l l o wrte i panE gih i n n li n l . t s () B ad n TTa p n eg 2 0 ) H ui i 1 M. o rma ,.rp e b r (0 6, e r t A sc fr re aa tr pi zt n i S p ot o Fe P rmee O t ai w t u p r mi o h V co Mahn sWC I 0 61 3 -3 4 etr c ie, C 2 0 ,3 71 4 .
  • Team Member Names: Jason Harris, Stephanie Winston, Sarah Digdon School of Information Management, Faculty of Management, Dalhousie University, Halifax, CanadaIntroducing Reference Rundown Integration Points User workflowWe are proposing a new SciVerse RR will integrate into ScienceDirect full text and the RR automatically loads the article‟s references andapplication titled Reference Rundown Hub Home Page. begins comparing them with references in other articles. On ScienceDirect platforms, the application will run in Clicking on the app expands the display to showA networking application that provides the right hand sidebar applications space. The small references and “shared reference” articles. Sorting andlinks between articles based on their referenced RR window will expand to allow users more sorting refining options are provided and users can click directlysources. Users will have an interface that allows and refining on a desired article within RR‟s reference list to openthem to: view, sort, and save articles that share one, options. another window with the chosen article in full text form.or more, cited sources with a selected article. Image of Science RR„s Direct view Expanded work flow.Value for Users integration into the HubThe RR application will help researchers find articles Home Pagethat share source material with a resource they have will allowalready discovered. Connecting articles based on users directsimilar sources can help users: access to the application.• Identify the authoritative voice on the topic• Find historical research that theories are based upon The Hub• Locate opposing views on a select topic Homepage• Find articles with different keywords from original will feature an advance search box permitting users to research locate the article they wish to have compared by RR.Once RR has generated the list of articles that sharereferences, users can modify the results to suit their Data Similar Applicationsneeds by eliminating The SciVerse MostCited application offers similar The RR application will need to access an index ofIrrelevant matches reference comparison services to RR. However, this app references that belong to individual articles. Scopusor specifying only operates on the Scopus platform, whereas RR will be the primary data source for this index as itdate ranges. functions on ScienceDirect Full Text and the Hub boasts “24.5 million records with references back to Homepage platforms. RR also provides more options for 1996” (SciVerse, n.d. par. 3). result refinement than currently offered by Most Cited. Reference: SciVerse. (n.d.) What does it cover. Retrieved from http://www.info.sciverse.com/scopus/scopus-in-detail/facts
  • Social Media Lab – Elsevier WebApp Concept Design Competition Map-a-Researcher Melissa Goertzen, Heather MacFadyen, Melissa Scanlan School of Information Management, Dalhousie University, Halifax, Canada Introduction User Workflow APIs Map-a-Researcher is a networking and visualization • Google Maps API Premier tool that harvests author names from the •Twitter REST API ScienceDirect database and plots the geographic •LinkedIn People Search API location of their affiliated institutions using Google Maps. The goal is to help graduate students connect •LinkedIn Get Network Updates and Statistics API with scholars who may supervise thesis projects. •LinkedIn Profile API. Data 1. SciVerse: provides author data such as geographic location, e-mail address, and phone number 2. Scopus: provides scholar’s h-index and citations 3. Google Maps: provides map data through Google Maps Premier API 4. LinkedIn/Twitter: provides most recent status updates and profile photo Figure 4. Contact tab: displays authors’ institutional affiliation, email address, phone number, and the most recent status updates posted in Integration Points LinkedIn and Twitter. The application will run within ScienceDirect Full Text Article Search. Figure 5. Citation tab: displays ranking information from the h index and the top 10 articles ranked by h factor. Figure 1. Visualization of how the Map-a-Researcher application is integrated into ScienceDirect. An API connection between SciVerse and Google Maps will integrate author and geospatial data Figure 6. Save tab: displays options for saving an author’s profile. For Further Information Figure 2. Visualization of how article selections are passed to the Map-a-Researcher application More information on this and related projects can be when the user clicks the Search button. Figure 3. Description of how integration is achieved and data flows between users, ScienceDirect, and API connections. found at www.SocialMediaLab.caPOSTER TEMPLATE BY:www.PosterPresentations.com
  • Team Member Names: Abdulhady Habash, Jodi Hunt, Andrea Prada Serrano, Maher Shawli School of Information Management, Faculty of Management, Dalhousie University, Halifax, Canada 1. When the users add the application, it will immediately open on the right handThe application will essentially work like a newspaper side of the screen. If the users are interested to post a comment, they will clickcomment section. It will be specific to each article. on the “New Comment” button. When somebody has already posted commentsThe user will be able to subscribe to comments and in this article, the user will be able to see them below the “New Comment”questions regarding that article. They will also be able button.to post comments for other researchers or even postcomments or questions for the authors, which in turnwill e-mail the post to the authors. When the author(s) 2. This is the form that appears after the user clicks on the “New Comment”reply, the response will be visible for everyone to see. button. Once the user fills in all the fields, they will hit Send.For anyone that has subscribed to posts for that article,they will receive an e-mail when a new response orcomment has been added.A typical user of this application would be one who may 3. When the user clicks on the comment or reply, a pop-up box will appear. In thishave further questions about an article and want to get pop-up box the user can read all the comments about that article and/or add ain contact with the author or other researchers. This response.application will avoid any duplicate e-mails that theauthor may receive regarding the same topic in an Acknowledgementsarticle, as others will be able to see the author’s reply. The authors acknowledge the Elsevier Web App Concept Design Competition for funding the printing of this poster. Special thanks to Dr. Anatoliy Gruzd for his support.
  • Team Member Names: Heather Buchanan, Jackie Phinney, Katie MacDonald School of Information Management, LIS, Dalhousie University, Halifax, Canada October 2011 P Designed with the user in mind Science Direct User Feedback “I … got a paper that was 9 months old, and Integration Point Current Month according to Scopus has 2 citations, and according to SD had zero citations. I do not see how such a paper can be described as HOT!” The Problem: Application Value How to deliver the newest In a fact-paced field , it allows users to stay regularly informed and hottest articles quickly Provides a forum for discussion among usersDesign Option #1 and easily to the user Keeps users up-to-date on new developmentsDesign Option #2 that may effect their own research/work The Solution Potential for educational value, in conjunction with Pharmacology subject guides Focus: User engagement & interaction Simple sorting mechanisms based on user Easy sorting by user needs/interests 0 needs Most Cited As determined by the Most Accessed number of comments Most Downloaded generated by a given APIs include: Most Buzzworthy article SciVerse Content API Star ratings for each article to reflect users’ interest in or Scopus RSS Feed agreement with individual articles Fivestar by Drupal Direct links to various forms of social media for easy QR Code sharing Facebook Updated on a monthly basis, staying current and relevant Twitter
  • Team Member Names: Kathleen MacKeigan, Douglas Seaman, Jessica Stewart School of Information Management, Faculty of Management, Dalhousie University, Halifax, CanadaWhat It Is User Workflow Similar ApplicationsThe Bibliographic Visualizer provides an interactive Cite Me!visual display that shows bibliographic 1. Select an article of interest from within your search Uses the immense amount of metadata on hand toassociations between academic articles. The results. Launch the visualizer and the selected article will be copy and paste a citation in a number of citationapplication is intended for: researchers building a displayed in the middle of the screen. Articles that cite your formats and create a bibliographic aide. Yet whilebibliography for a literature review; professors initial choice will appear, branching out from the original. Cite Me! views reference material as an end incompiling a reading list; students searching for itself, our application’s use of metadata wouldinfluential articles; or, anyone else who is allow the researcher to expand their search.conducting academic research. 2. Display the metadata of any article within the visualization by hovering over itIntegration Points with your mouse. Save this metadata by clicking the checkbox at the upper-right Prolific Authors corner. Move the new article to the center by left-clicking on it. Offers users the ability to identify authorities in a• ScienceDirect: the toolbar will appear on the given field. Our application allows users to identify right hand side of the full-text article page. those authors who defy traditional notions of• Scopus: the toolbar willv appear on the right hand disciplines, either by having their work cited in side of the abstract page. articles whose subject is traditionally considered• Hub Search results page: the toolbar will appear outside of the cited author’s domain, or through directly under the selected record. their own use of unorthodox ideas and resources. 3. Explore deeper 4. Filter branched 5. Export the list of levels of bibliographic articles by the following selected articles intoValue for Users Search Term Timeline records until you have parameters: RefWorks or into a Word Provides telling research and publicationThe Bibliographic Visualizer allows users to: found your relevant  Author(s) document. This list will information for any period a user is interested in• see the cross-disciplinary nature of articles; articles, or until there  Date range include a given article’s investigating. The user of the Search Term• quickly gage article popularity and influence by are no more articles to  Journal title(s) metadata in addition to Timeline would be someone who was interested in seeing how many times an article has been explore. You can return  Subject the bibliographic record the popularity, or lack thereof, of a given subject, cited in other articles; to previous level any level in which the article our application’s user would be predominantly area(s)/discipline(s)• find out who is publishing most extensively in a begin with an article and follow the transmission time by pressing the  Minimum number of was retrieved. particular subject area; and reinterpretation of the idea, rather than its Back button. citations• see various interconnections between particular popularity. journals, authors, and articles. References: Elsevier B.V. (2011). Cite Me! Retrieved October 5, 2011 from http://www.applications.sciverse.com/action/appDetail/298329?zone=main&pageOrigin=appGallery&activity=display; Acknowledgements: The Bibliographic Visualizer group would like to thank the organizing committee of the Elsevier WebApp Elsevier B.V. (2011). Prolific Authors. Retrieved October 5, 2011 from http://www.applications.sciverse.com/action/appDetail/292667?zone=main&pageOrigin=appGallery&activity=display; Elsevier v Concept Design Competition: Dr. Anatoliy Gruzd, Vishal Gupta, Philip Mai, and Melissa Anez and Jennifer Yurchak. We would also B.V. (2011). Search Term Timeline. Retrieved October 6, 2011 from http://www.applications.sciverse.com/action/appDetail/292691?zone=main like to acknowledge the support of the Dalhousie Social Media Lab and the Dalhousie School of Information Management.
  • Social Media Lab – Elsevier WebAPP Concept Design Competition SciGuide Colleen Delany, Samantha Dutka, and Lora Hamilton The SciGuide Vision Jenny’s university’s live chat is available using the WorldCatWe envision SciGuide as an application API. She IM’s her librarian foraggregator that brings together multiple SciGuide: Search ScienceDirect help getting started.applications on one source page. Ask a Librarian Jenny’s Journey The librarian recommends•Our target population is first year the APA reference guide and research webinars university students and other first time embedded as E-Resources. researchers who do not have experience E-Resources with SciVerse or the research process. SciVerse YouTube Channel Jenny notices key words for•Let’s follow Jenny, a first-year biology Purdue Online Writing Lab (OWL) subsequent searches integrated major, on her journey through the search via APIs: • ScienceDirect controlled process using SciGuide. vocabulary Integrated SciVerse apps: •Natural language thesaurus • CiteMe! generates copy/pasteable citations SciGuide: Article Record SciGuide: Search Results • Most Downloaded shows CiteMe! popular articles from the Related key words same journal ScienceDirect Big Huge Thesaurus Get it at your library Most Downloaded The ScienceDirect search results are paralleled by the WorldCat results available at Jenny’s university, using the WorldCat API.
  • Problem Implementation• Collaboration among researchers is a must• There are a lot of tools and • Social media technology in ways to collaborate and SciVerse share • MyHub allows you to create• Which one do I use? a team which can involve anyone with an account on• Which one are my SciVerse colleagues using? • Team members can• What is the best tool for recommend items to the collaborating on research team, chat with other team on Elsevier platform ? members • Create chat API By: Jude Abbey , MEC • Use existing user management API • Use / add to exiting item metadata Victor Itiveh, MEC
  • o Enter SciVerse, login or create an account. WORKFLOW Concept by: Elinor Crosby o Click on the “Add” button from the My Applications bar, search for the Lending Library Kristy McGill Tammy Whynot application. o Install, then select a default citation style (APA, ALA, or Chicago). LENDING LIBRARY o Browse articles in Save to bookshelf Science Direct or Scopus. VALUE o Create & store citations in APA, ALA, or Chicago formats & gather them into bibliographies. o The Lending Library o Share citations or a bibliography via LinkedIn. icon will appear in the o One click to save, one click to share. toolbar or beside each o Access articles via a traditional folder and list article view, or through an innovative visual bookshelf o Click to “add” the article view. . to the personal bookshelf SHARE o Export citations and bibliographies to share o View articles on the your research interests bookshelf in a list with colleagues and alphabetically, by date students. added to collection o Or by viewing a SciVerse Information & Knowledge Management Information Systems Management “picture” of the first page of the article Research Subject - Date o Export formatted CITE o Create folders for better organization. citations and entire o Drag &drop to rearrange, whether in list view or bibliographies to an bookshelf view. Open Office document. Research Topic 1 Research Topic 2 ReferencesElsevier B.V. (2011a). Cite Me! Overview. Retrieved from Research Topic 3http://www.applications.sciverse.com/action/appDetail/298329?zone=main&pageOrigin=appGallery&activity=display Research Topic 3Elsevier B.V. (2011a). SciverNote Overview Retrieved fromhttp://www.applications.sciverse.com/action/appDetail/293244?zone=main&pageOrigin=home&activity=display
  • Social Media Lab – Elsevier WebApp Concept Design Competition On-Screen Figure Digitizer Stanley Selig, Colin O’Flynn - Faculty of Engineering , Dalhousie University, Halifax, CanadaApplication Description IntegrationThe On-Screen Figure Digitizer allows the reader to The application functions best if integrated into theperform operations on figures located within a paper. “Full-size image” window, but this would require minorWhen a figure is clicked on, a menu will allow the reader changes to the API by Elsevier. If this is not possible theto perform these tasks. On-screen measurements and a application can be integrated into the ‘My Applications’graph digitizer are the key functions envisioned. For any window, and the user drags-and-drops an image file tofigure with a scale, the measurement tools will enable the application.a reader to take accurate measurements directly off thefigure, including distance and area. The graph digitizerwill take reader-input axes and scales which will be su-perimposed on the graph, and export data points fromcurves in the figure.Value for UsersThis application can be used by a very diverse group ofresearchers including those in the fields of medicine,engineering, economics, and natural sciences. The mea-surement tools can help readers make quick referencemeasurements while reading a paper without having to Graph Dataopen external programs; for example, identifying cell Easily extract data from graphs in published papers.wall thicknesses on SEM images. Users that read papers Click two reference points on each axis, enter the val-that include maps or regions would benefit from this ap- ues of those reference points, and click points within theplication by being able to calculate areas on-the-fly based graph to get values instantly. Both linear and logarith-on a scale present on a figure. The graph digitizer would mic scales are supported.be helpful by allowing the user to export manipulable Since the application appears in the “Full-size image”versions of graphs found in a paper, and thereby let the window, you can open graphs from different papers anduser determine important information, such as precise Bounded Area On-Screen Ruler compare results, as one application follows each graph.data points, slopes, and functions/trendlines. This can Based on a reference ruler included in a figure, measure Based on a reference ruler included in a photograph Includes option to automatically recognize graph databe useful to a reader who would like to use the paper’s the area inside a polygon you create. Measure areas in or drawing, take any additional measurements you re- and download for further processing.data for further analysis. maps or under graphs. quire.
  • Social Media Lab – Elsevier WebApp Concept Design Competition Tree of Knowledge: Bringing Elsevier’s Tree to Life Marek Lipczak1, Tomasz Niewiarowski1 and Fatemeh Riahi1 1 Faculty of Computer Science, Dalhousie University, Halifax, CanadaProblem: exploring scientific literature Tree of Knowledge System architecture Old approach: literature accessed through text-based search Query: support vector machines Pros: standard approach to textual content access Cons: one dimensional list of results focuses on text and is missing important aspects of scientific literature Our solution: tree based visualization of retrieved articles Natural and readable representation of knowledge formulation process Target user - new graduate student Tree generation module Features Input: directed, weighted graph of documents (citation links Representation of various article features (recency, and content similarity) relations, topic, importance) Easy access to founding articles and related articles Division into research sub-areasTree visualization (realistic abstract) Processing: graph based clustering algorithm Minimum spanning tree/graph coarsening Adapted MCL (Markov CLuster algorithm) basic solution experimental solutionInterface Bottom of the tree represents a set of founding articles for the area (most cited articles) The branches split into more specific sub-areas Labeling: extract most important keywords for each branch Documents represented as leaves Visualization: open-source JavaScript libraries Tree features Leaves raphael.js d3.js processing.js size – importance (number of citations) colour, position – recency (publication year) Branches thickness – number of related articles labels – keywords extracted from articles Navigation References: sources and inspirations SciVerse Hub search interface: http://www.hub.sciverse.com/action/home zooming, panning Figures: Citation graph (Action Science Explorer): http://www.cs.umd.edu/hcil/ase/ Minimum spanning tree: http://en.wikipedia.org/wiki/Minimum spanning tree hovering in search panel – matching between MCL algorithm figure: http://micans.org/mcl/ Visualization accessed through the search interface documents in the search results list and the tree Inspirations: Line based tree visualization: http://en.wikipedia.org/wiki/Iranian languages Obama’s tree: http://weblog.bocoup.com/newsweek-raphaeljs-career-tree-visualization-raphaeljs-burst Double-click expands the tree to full screen hovering in tree panel – expanding information about Code: MCL algorithm: http://micans.org/mcl/ raphael.js: http://raphaeljs.com/ d3.js: http://mbostock.github.com/d3/ Hovering over a document/branch reveals detailed description leaves or branches processing.js: http://processingjs.org/Machine Learning and Networked Information Spaces group http://www.cs.dal.ca/˜lipczak/treeOfKnowledge.php lipczak@cs.dal.ca
  • Team Member Names: Seyednaser Nourashrafeddin, Armin Sajadi Faculty of Computer Science, Dalhousie University, Halifax, Canada Primary  Purpose  of  Applica1on   Target  User   Visualiza1on  •  Grouping  the  search  results  into  some  categories •  The  target  user  for  this  applica7on  is  a  newbie  in   Whenever  the  user  hover  the  mouse  over  a    in  a  way  that    each  category  represents  a  field  the  field  defined  by  the  query.     paper,  the  tool  shows  a  heat  map  of  the  paper    exis7ng  in  the  retrieved  paper,s  like  what  a for  corresponding  query  and  offering  keywords.   •  This  user  most  probably  does  not  know  how  to  survey  paper  aims  to  do.    formulate  his  query,  or  where  the  boundaries  of   Besides,  the  tools  shows  the  subclusters  to  •  Offering  similar  keywords  to  the  user’s  query.    the  field  are.  This  applica7on  gives  her  some  sets which  the  paper  belongs  to  and  their    of  keywords  and  also  papers  related  to  those corresponding  heat  maps.  •  Visualizing  the  relevance  of  each  paper  to  the  submiDed  query  using    heat  map.    keywords.       By  comparing  the  heat  map  of  the  paper  and  the   average  heat  maps  of  each  subcluster,  the  user  is  •  Visualizing  the  most  similar  papers  to  each able  to  select  similar  papers.    selected  paper  in  different  categories  using  heat  maps  of  subclusters.   Workflow     Only  the  submiDed  and  offering  keywords  are   shown  in  heatmaps.   1-­‐The  user  submits  a  query.   2-­‐The  search  engine  returns  some  papers  regarding Query keywords + Offered Keywords  the  submiDed  query.   Heat  Maps   3-­‐The  tool  extracts  7tles  and  abstracts  of  the  first pi keywordj  100  papers.   4-­‐The  tool  applies  a  co-­‐clustering  algorithms  to  find  clusters  of  papers  and  keywords.   5-­‐The  tool  offers  some  other  keywords  for  each paperi  submiDed  keyword  from  the  sub-­‐clusters pi  extracted  by  the  Co-­‐Clusterer.   pi Sub-Cluster 1 Sub-Cluster 2 query query Offered   KNN   pi Pale Green: least relevant Keywords   Search Engine Retrieved  Papers   {Cosine  Similarity}   {Title,  Abstract}   Keyword  Sub-­‐ Dark Green: less relevant clusters   Black : Relevant Co-­‐Clustering   Dark Red: More relevant query pi Heat  Maps   Pale Red: Most Relevant Paper   Sub-­‐clusters   Sub-Cluster 3 Sub-Cluster 4
  • Team Member Names: Laura Covert, Jennifer Gibbons-Bishop, Norma Livingstone, Whitney Spencer * School & Faculty, SIMS, Dalhousie University, Halifax, CanadaIntroduc)on   APIs A graphic interface that is responsive to user   Export to PDF input. The interface will also need to be able to   hGp://www.programmablewegenerate a image of the representa9on of the   b.com/api/pdfcrowd recalled bibliographic material  Data  Our applica9on will draw primarily  upon bibliographic records, as well  as frequency sta9s9cs, as a means  to rank the usefulness of the  bibliography.   Integra)on Points  ScienceDirect Full Text Ar9cle,  Scopus Abstract, Hub Search, and  Hub Home page.  Descrip)on User workflow  The user will be able to modify the graphic  Similar Applica)ons BiblioWeb begins like any normal search. Once  interface with tools that allow you to  Cite Me!, Co‐Author Network, Co‐an ar9cle is selected, the output will appear as  manipulate and customize the web of records  Author Visualizer, Scriver Note a visual representa9on of the recalled ar9cles,  for op9mal efficiency. Op9ons will include  from SciVerse, and SciVerse Scopus showing their connec9ons to other  altering or adding to the web as your need, bibliographic records.   and customize colour arrangements. 
  • Team Member Names: Ariel Kleber, Philip Taber, Kae Yee Tsai School of Information Management, Dalhousie University, Halifax, CanadaIntroduction Featuresj.rate is a sleek and simple research aid that will help students • By providing comparative ratings between the selectedfocus their research on the most significant journals in their field. article and journal, j.rate underlines the academic standing ofj.rate offers students the next step in research, providing a the journal as a viable information sourcerecognizable interface that utilizes social media tools such as • Journal rating fields including its standing in SCImagoFacebook in conjunction with Scopus and Elsevier data. Journal Rank, the H index, total citations by year, totalEmbedded in each search window the application provides downloads, reviews, and more.comprehensive journal ratings and rankings next to eachindividual article. • Specific data including article citations, article downloads, article reviews, andThis tool will help students find the most reliable academic recommendations are the indicators that canresources available to them. help the student determine the overall critical reception of the journal • The easy-to-use interface automatically updates with each new search • A mouse-over glossary explains the criteria of each separate rating system • j.rate also includes embedded links to the original data sources, such as the SCImago source page as well as explanations of the numerical rankings • The downloadable application would extend this service beyond Elsevier databases, catering to the diverse research needs of the broad academic population. References: http://www.sciencedirect.com/science/article/pii/S0960982295001369 http://www.scimagojr.com/journalrank.php
  • R e s e a rc h Ne w s Wa tc h by Doyle Lahey, Mark McHale & Daniel Teed FUTURE APP TO LINK ACTIVE AND PASSIVE SEARCH THE NEW WITH THE NEWS First, the user would either conduct a basic orResearch News Watch will provide subscribers advanced search depending on how specificwith the ability to conduct a search of a current their topic is and how much prior research theynews item and find journal articles in the have already completed. If they are justSciverse databases that relate to the original beginning their research endeavour, they wouldsearch topic. This will help facilitate a enter either a topic, subject term, or field intoconnection between previous academic the search box. In addition to the articles thatresearch and mainstream news media, which RESEARCH NEWS WATCH would normally appear in a SciVerse search, awill help provide a broader comprehension of box in the sidebar would list news items relatednews topics. In addition, users may also identify AN ACADEMIC ARTICLE OF SOME INTEREST to their search. When a user clicks on an articlenews articles that are related to any given to display the full record, the box would alsoacademic publication. RELATED NEWS HEADLINE- In recent news, a related news item has appeared the is relevant to this ... appear on that screen with updated results A DIFFERENT NEWS ARTICLE - based on the particular item that they have Similar to the article above, this media outlet has recently printed a similar... chosen. BREAKING NEWS: ANOTHER CONTEXT AND IMMEDIACY HEADLINE - There are unconfirmed reports that another relevant news article has... To provide the opposite service, linking specificThis application would be particularly helpful MORE news items to recent journal articles, a cyclingfor students as it would allow them to see the news feed would display news headlines andconnections between academics and issues in several related journal articles for twentythe world. Furthermore, it will help students seconds at a time. This feed would be locatedconduct research about a current topic that USES EXISTING METADATA so that it sits in either the top or bottom cornerthey may be writing about, along with As the core function of the application is to identify the of the site, providing easy access without beingprofessors who want their students to relevance of academic articles to news articles, and vice distractingly prominent.contextualize what they are learning in the versa, it would have to be able to access such data asclassroom. It would also allow any user of would be necessary to determine the level relevance.SciVerse to pursue their curiosities, thus making The application would also need to be developed toSciVerse more engaging. This application would use the data of various news organizations, each ofalso be of use for media professionals by which could have its own metadata schemes. Thisproviding properly researched information metadata would include keywords and subject terms,related to current topics. Essentially, this feature as well as dates of publication. Data on the number ofwould allow for different perspectives and citations of any given journal article could also help toencourage a more balanced presentation of determine the relevance.news topics.