The life sciences domain has been one of the early adopters of Open Data Initiative and a considerable portion of the Linked Open Data cloud is comprised of datasets from Life Sciences Linked Open Data (LSLOD). This deluge of biomedical data and active research over the past decade resulted in the flux of scientific publications in this domain. PubMed resource provides access to MEDLINE, NLM's database of citations and abstracts in the biomedical domain. PubMed Central provides links to full-text articles along with publisher web sites, and other related resources. In this paper we present PubMed Visual Interface (PrEVIEw)a web based application to access information related to publication, research topic, author and institute through a visual interface. PrEVIEw not only provides useful information e.g. research topic of interest, research collaboration at personal or institute level etc, for the biomedical research community but also helpful for the working Data Scientist. We also evaluate the usability of our system by using the standard system usability scale as well as a custom questionnaire, particularly designed to test the usability of the interface. Our overall usability score of 83.69 suggests that web based interface is easy to learn, consistent, and adequate for frequent use.
The deluge of biomedical data in the last few years, partially caused by the
advent of high-throughput gene sequencing technologies [
The remaining part of this paper is organised as follows: we highlight the related work in section 2. We introduce our methodology and PrEVIEw salient features in section 3. Later we present the usage scenario in section 4. Subsequently, we present an evaluation of our approach in section 5. We finally conclude the paper. 2
AMiner by Tang J et. al creates semantic-based profile for each researcher in ored
to built researcher social network.[
TPX [
The table 1 shows that PrEVIEw is not limited for searching particular topic,
not confined for searching articles for specific time frame, or specific database of
pubmed neither for the social networking purposes. Furthermore tabular display,
can be difficult to pull off linked information while concept map approach [
Our methodology consists of two steps namely: 1) result retrieval using Rest API’s and 2) building visual interface .
Result Retrieval Using Rest API’s REST (Representational State Transfer)
is a communication approach that is often used in the development of Web
services6 and involves reading any Web page that contains an XML file. PrEVIEw
6 http://searchsoa.techtarget.com/definition/REST
make use of REST APIs7 provided by PubMed and overall result retrieval works
as follow: (1) User search a topic, institute, title of the publication or author and
send the request, (2) Rest API’s fetch data in the form of XML, (3) XML parser
with data manipulator creates JSON file and transfer to presentation layer.
Building Visual Interface The analysis of large numbers of biomedical
publications, in the form of graphical representation made it possible for researchers
to extract the relevant publications for future collaboration. We chose the
concept map approach [
Technologies PrEVIEw is browser-based client application that provides a flexible front-end for searching. To build this application variety of web technologies are used including HTML5, CSS, JavaScript, JQuery8, Rest API’s 9, Java Servlet, SVG10, AJAX11 and JSON12.
Availability The PrEVIEw application can be accessed at http://srvgal78. deri.ie/PrEVIEw/. 3.1
Fig. 1: PrEVIEw Architecture: Data, Business and Presentation layer.
all the relevant publications is retrieved at the Data Layer and corresponding
JSON is generated at the Business Layer. At the Presentation Layer,
graphical and visual representation of PId’s (PubMed Id’s), authors and institute are
displayed as clusters. Complete metadata description of publications can be
accessible by clicking on PId’s . For the second challenge, the classification, we
adopt the technique that relies on publication mapping, a web-based client
application PrEVIEw is developed. Publications are mapped in the form of clusters
classifying author, topic, title and institute separately. Authors and institutes
with remarkable contribution in regard to particular topic search, represented as
larger node. Additionally, the publications can be downloaded. We applied two
mapping techniques to this application: force directed placement [
Recognising the value of research networking, two example scenarios are discussed to demonstrate the use of PrEVIEw. First use-case visualises PubMed ID’s and authors as the outcome from topic/title query. For second use-case, the institutional outcomes have been used. The combination of these two usecase enables establishing research networks and collaborations at authors and institution level. Details of these two outcomes have been discussed below. Highly cited publications with the author list having paper on ”cancer” research The step-by-step approach is discussed as follow (Fig. 2): 1. The first step is to specify the selection of the corresponding text. User search ”cancer” while selecting Topic from the drop-down menu. (window A). 2. The visualisation of top 100 highly cited papers on cancer can be seen in (window B) where user can explore them using PubMed Id’s and and description about the full paper can be explored (window D). 3. List of all authors (from top 100 cited papers) working on cancer domain is shown in (window C) where a mouse-hover further display the ”author’s name”. 4. Selected author contribution in cancer research (window E) provide user the ability to download all the relevant papers in cancer domain.
Authors collaborations at international level increase citations of research manuscripts. One of the novel advancement in the mode represented in Figure 2 is establishing authors research networks and collaborations on a focused domain.
Research Institutes working on ”cancer research”, and their collaborations with other institutes The step-by-step approach is discussed as follows(Fig. 3): 1. The first step is to specify the selection with the corresponding text. In this case ”cancer” while selecting ”Topic” from ”drop-down menu” (window A). 2. The visualisation of highly cited cancer papers can be seen (window B) where user can explore them using PubMed Id’s. 3. List of all institutes working on ”cancer research” are shown in (window C) where user is able to explore any of them by selection. 4. Selected institute collaboration with other institutes in cancer research is shown in publication metadata, (window D).
The exponential growth in international collaboration on focused scientific research questions shows the novelty in the mode represented in Figure 3. Institutional collaboration enables shared learning, new research opportunities, establishing new research projects, joint applications for funds, and technology transfer. Findings from these different kinds of networks can be used in many ways. Collaboration networks in terms of highly cited authors , institutions, and countries are highlighted in the publication set. Semantic networks can identify various research directions the focused concepts. Further more publication citation networks can be used to quantify the citation impact in research directions and disciplines. 5
The goal of our evaluation is to quantify the usability and usefulness of
PrEVIEw graphical interface. We evaluate the usability of the interface by using the
standard System Usability Scale (SUS) [
In this section, we explain the SUS questionnaire13 results. This survey is more
general and applicable to any system to measure the usability. The SUS is a
simple, low-cost, reliable 10 item scale that can be used for global assessments
of systems usability[
As of 3rd May 2016, 22 users14 including researchers and engineers in Semantic Web participated in survey. According to SUS, we achieved a mean usability score of 83.69 indicating a high level of usability according to the SUS score. The average scores (out of 5) for each survey question along with standard deviation is shown in Figure 4.
The responses to question 1 (average score to question 4.2 = ± 0.72) suggests that PrEVIEw is adequate for frequent use. The responses to question 3 indicates that PrEVIEw is easy to use (average score 4.5 ± 0.51) and the responses to question 7 (average score 4.4 ± 0.59) suggests that most people would learn to use this system very quickly. However, the slightly higher standard deviation to question 9 (standard deviation = ± 0.74) and question 10 (standard deviation 13 SUS survey can found at: https://goo.gl/hkuMDM 14 Users from AKSW, University of Leipzig and INSIGHT Centre, National University of Ireland, Galway. Summary of the responses can be found at: https://goo.gl/ iKaZjQ = ± 0.39) suggest that we may need a user manual to explain the different functionalists provided by the PrEVIEw interface. This survey15 was particularly designed to measure the usability and usefulness of the different functionalists provided by PrEVIEw. In particular, we asked users to use our system and share their experience through question 9 and question 10. As of 3rd May 2016, 13 researchers including Computer Scientist16 and Bioinformaticians have participated in survey. The average scores (out of 5 with 1 meaning strongly disagree and 5 meaning strongly agree) for each survey question along with standard deviation is shown in Figure 5. The average scores to question 10 (i.e., 4.45 ± 0.73), show that most of the users feel confidence to use the system and need not to learn much about the PrEVIEw before using. The responses to question 2 (average score = 4.07 ± 0.64) suggest that exploring highly cited publications using PubId selection is easy in PrEVIEw. A slightly lower scores to question 7 (average score = 4.38 ± 0.51) suggests that we need to further improve the user experience with visualisation components of the PrEVIEw. 15 Custom survey can be found at: https://goo.gl/syZzAM 16 Summary of the responses can be found at: https://goo.gl/JKogDQ 6
In this paper we introduce PrEVIEw as an online visual and graphical interface for searching PubMed resource. We evaluate our approach and usability of our system using the standard system usability scale as well as through domain experts. Our preliminary analysis and evaluation revels the overall usability score of 83.69, concluding PrEVIEw an interface, easy to learn and help users accessing PubMed resource intuitively. As a future work we aim to extend PrEVIEw with Faceted browsing and also provide visualisation at entity level e.g, Genes and Molecules where the search criteria retrieve these entities. Current work visualise top 100 cited results and in future we aim to visualise all retrieved results.
This research has been supported in part by Science Foundation Ireland under Grant Number SFI/12/RC/2289.