Research Assistant/PhD Candidate, University of Alberta
A PubMed Citations Recommender System for Real-Time Chat
In our paper, we give details about the development and evaluation of a recommender system targeting online health discussions. The recommender system, called PubMedReco, analyzes conversations going on in real-time in chat rooms or forums. PubMedReco then shows links to medical articles that are relevant to the conversations. The medical articles are linked from PubMed, a trusted and popular database of over 23 million medical citations. Our tool is especially useful for health chat rooms and forums. It uses machine learning to determine the most relevant medical keywords in the conversations and then automates the process of searching and finding related medical articles on PubMed.
Abstract: Prevailing health care structures and cultures restrict intraprofessional communication, inhibiting knowledge dissemination and impacting the translation of research into practice. Virtual communities may facilitate professional networking and knowledge sharing in and between health care disciplines.This study aimed to review the literature on the use of social media by health care professionals in developing virtual communities that facilitate professional networking, knowledge sharing, and evidence-informed practice.An integrative literature review was conducted to identify research published between 1990 and 2015. Search strategies sourced electronic databases (PubMed, CINAHL), snowball references, and tables of contents of 3 journals. Papers that evaluated social media use by health care professionals (unless within an education framework) using any research design (except for research protocols or narrative reviews) were included. Standardized data extraction and quality assessment tools were used.Overall, 72 studies were included: 44 qualitative (including 2 ethnographies, 26 qualitative descriptive, and 1 Q-sort) and 20 mixed-methods studies, and 8 literature reviews. The most common methods of data collection were Web-based observation (n=39), surveys (n=23), interviews (n=11), focus groups (n=2), and diaries (n=1). Study quality was mixed. Social media studied included Listservs (n=22), Twitter (n=18), general social media (n=17), discussion forums (n=7), Web 2.0 (n=3), virtual community of practice (n=3), wiki (n=1), and Facebook (n=1). A range of health care professionals were sampled in the studies, including physicians (n=24), nurses (n=15), allied health professionals (n=14), followed by health care professionals in general (n=8), a multidisciplinary clinical specialty area (n=9), and midwives (n=2). Of 36 virtual communities, 31 were monodiscipline for a discrete clinical specialty. Population uptake by the target group ranged from 1.6% to 29% (n=4). Evaluation using related theories of "planned behavior" and the "technology acceptance model" (n=3) suggests that social media use is mediated by an individual's positive attitude toward and accessibility of the media, which is reinforced by credible peers. The most common reason to establish a virtual community was to create a forum where relevant specialty knowledge could be shared and professional issues discussed (n=17). Most members demonstrated low posting behaviors but more frequent reading or accessing behaviors. The most common Web-based activity was request for and supply of specialty-specific clinical information. This knowledge sharing is facilitated by a Web-based culture of collectivism, reciprocity, and a respectful noncompetitive environment. Findings suggest that health care professionals view virtual communities as valuable knowledge portals for sourcing clinically relevant and quality information that enables them to make more informed practice decisions.There is emerging evidence that health care professionals use social media to develop virtual communities to share domain knowledge. These virtual communities, however, currently reflect tribal behaviors of clinicians that may continue to limit knowledge sharing. Further research is required to evaluate the effects of social media on knowledge distribution in clinical practice and importantly whether patient outcomes are significantly improved.
Pub.: 23 Jun '16, Pinned: 28 Jun '17
Abstract: The use of social media has become very instinctive to many. It has become part of a daily routine. Enhanced communication, liberated expressions of self, becoming updated with all the trends and news, and marketing promotion are only some of the reasons why most people use social media. Health care providers including physicians should take advantage of these platforms for professional purposes. Social media extends far beyond the famous platforms such as Facebook, Twitter, Pinterest, and Instagram, used mostly for social connections. There are sites dedicated to serve professionals, for example, LinkedIn, or even physician-specific forums such as Sermo. The physical medicine and rehabilitation community has a forum (Phyzforum) created by the American Academy of Physical Medicine and Rehabilitation to share questions, comments, and ideas. Moreover, there are broadcast media (Podcast) and blogging sites (WordPress) used by many physiatrists. Surveys show that physicians actively use an average of 2-4 hours of professional-leaning networking sites per week; for example, 44% of physicians use Sermo and 42% use LinkedIn. The participation also extends to more popular sites, with 40%, 25%, and 20% physician participation in YouTube, Blogging, and Twitter, respectively. There are numerous guidelines available for medical practitioners pertaining to professional use of social media. Strategies such as timing of postings and posting content as well as methods to maintain your online reputation are discussed. Various benefits and potential pitfalls with regards to social media use are also analyzed, including how to engage followers and addressing negative comments and reviews.
Pub.: 22 May '17, Pinned: 28 Jun '17
Abstract: Online health forums provide peer support for a range of medical conditions including life-threatening and terminal illnesses. Trust is an important component of peer-to-peer support, although relatively little is known about how trust forms within online health forums.The aim of this paper is to examine how trust develops and influences sharing among users of an online breast cancer forum.An interpretive qualitative approach was adopted. Data were collected from forum posts from 135 threads on 9 boards on the UK charity, Breast Cancer Care (BCC). Semistructured interviews were conducted with 14 BCC forum users. Both datasets were analyzed thematically using Braun and Clarke's approach and combined to triangulate analysis.Trust operates in 3 dimensions, structural, relational, and temporal, and these intersect with each other and do not operate in isolation. The structural dimension relates to how the affordances and formal rules of the site affected trust. The relational dimension refers to how trust was necessarily experienced in interactions with other forum users: it emerged within relationships and was a social phenomenon. The temporal dimension relates to how trust changed over time and was influenced by the length of time users spent on the forum.Trust is a process that changes over time and which is influenced by structural features of the forum, as well as informal but collectively understood relational interactions among forum users. The study provides a better understanding of how the intersecting structural, relational, and temporal aspects that support the development of trust facilitate sharing in online environments. These findings will help organizations developing online health forums.
Pub.: 26 May '17, Pinned: 28 Jun '17
Abstract: To investigate online self-help forums related to cannabis users who were searching for help on the Internet.We analyzed the content of 717 postings by 328 users in three online forums in terms of fields of interest and self-help mechanisms. Only English-language forums that were free of charge and without registration were investigated.The main self-help mechanisms were disclosure and symptoms, with relatively few posts concerning legal issues and social perceptions. The forums differed significantly in all fields of interest and self-help mechanisms except for social network and financial and vocational issues. Highly involved users more commonly posted on topics related to diagnosis, etiology/research, and provision of information and less commonly on those related to gratitude. Correlation analysis showed a moderate negative correlation between emotional support and illness-related aspects and between emotional support and exchange of information.Cannabis forums share similarities with other mental health forums. Posts differ according to user involvement and the specific orientation of the forum.The Internet offers a viable source of self-help and social support for cannabis users, which has potential clinical implications in terms of referring clients to specific forums.
Pub.: 13 Jun '17, Pinned: 28 Jun '17
Abstract: Children from families with a mental illness are at risk of developing negative health outcomes. Online interventions are a new way to offer support to these children. The present study utilized a website that had been developed to support Dutch youth who had a family member with a mental illness. The objective was to analyse monitored and unmonitored chatroom conversations among these young people, and specifically to compare supportive messages and self-disclosures of experiences. We electronically imported session transcripts of 34 chatroom conversations into the qualitative analysis software Atlas.ti. A content analysis was performed on 4252 messages from 22 female participants. A correlational analysis was then conducted to identify significant associations between sent and received supportive statements and disclosing statements. We found supporting comments in approximately 34% of the conversations and disclosures of problems in the home in approximately 15–18% of the messages. Participants made approximately twice as many disclosing statements and approximately half as many supportive statements in the monitored sessions compared to the unmonitored sessions. The number of disclosures that were sent was positively correlated with the amount of social support that was received. The number of disclosures sent was negatively correlated with the amount of social support that was sent, but only in the unmonitored sessions. Considering the greater reach of Internet interventions, online chatroom sessions might be provided as complementary to, or as an alternative to, face-to-face groups for supporting youth with a family member who has a mental illness.
Pub.: 01 Jun '17, Pinned: 28 Jun '17
Abstract: Key opinion leaders (KOLs) are people who can influence public opinion on a certain subject matter. In the field of medical and health informatics, it is critical to identify KOLs on various disease conditions. However, there have been very few studies on this topic.We aimed to develop a recommender system for identifying KOLs for any specific disease with health care data mining.We exploited an unsupervised aggregation approach for integrating various ranking features to identify doctors who have the potential to be KOLs on a range of diseases. We introduce the design, implementation, and deployment details of the recommender system. This system collects the professional footprints of doctors, such as papers in scientific journals, presentation activities, patient advocacy, and media exposure, and uses them as ranking features to identify KOLs.We collected the information of 2,381,750 doctors in China from 3,657,797 medical journal papers they published, together with their profiles, academic publications, and funding. The empirical results demonstrated that our system outperformed several benchmark systems by a significant margin. Moreover, we conducted a case study in a real-world system to verify the applicability of our proposed method.Our results show that doctors' profiles and their academic publications are key data sources for identifying KOLs in the field of medical and health informatics. Moreover, we deployed the recommender system and applied the data service to a recommender system of the China-based Internet technology company NetEase. Patients can obtain authority ranking lists of doctors with this system on any given disease.
Pub.: 09 Jul '16, Pinned: 28 Jun '17
Abstract: The Internet, and its popularity, continues to grow at an unprecedented pace. Watching videos online is very popular; it is estimated that 500 h of video are uploaded onto YouTube, a video-sharing service, every minute and that, by 2019, video formats will comprise more than 80% of Internet traffic. Health-related videos are very popular on YouTube, but their quality is always a matter of concern. One approach to enhancing the quality of online videos is to provide additional educational health content, such as websites, to support health consumers. This study investigates the feasibility of building a content-based recommender system that links health consumers to reputable health educational websites from MedlinePlus for a given health video from YouTube.The dataset for this study includes a collection of health-related videos and their available metadata. Semantic technologies (such as SNOMED-CT and Bio-ontology) were used to recommend health websites from MedlinePlus. A total of 26 healths professionals participated in evaluating 253 recommended links for a total of 53 videos about general health, hypertension, or diabetes. The relevance of the recommended health websites from MedlinePlus to the videos was measured using information retrieval metrics such as the normalized discounted cumulative gain and precision at K.The majority of websites recommended by our system for health videos were relevant, based on ratings by health professionals. The normalized discounted cumulative gain was between 46% and 90% for the different topics.Our study demonstrates the feasibility of using a semantic content-based recommender system to enrich YouTube health videos. Evaluation with end-users, in addition to healthcare professionals, will be required to identify the acceptance of these recommendations in a nonsimulated information-seeking context.
Pub.: 17 May '17, Pinned: 28 Jun '17
Abstract: Method, system, and computer program product for personalized medical content recommendation are provided. The method may comprise having a patient medical profile; having a concept or relationship relating to the patient medical profile; obtaining medical content relevant to the concept or relationship; determining a score of the concept or relationship to the patient medical profile; enhancing the score with a context score based on the values of attributes of the concept or relationship in the patient medical profile; enhancing the score with an additional knowledge score based on knowledge of the concept or relationship additional to the medical content; recommending medical content with respect to a patient medical profile based on the enhanced scores. The method may also include providing an explanation of a medical content recommendation based on a matching concept or relationship.
Pub.: 24 Sep '13, Pinned: 28 Jun '17
Abstract: Exploiting temporal context has been proved to be an effective approach to improve recommendation performance, as shown, e.g. in the Netflix Prize competition. Time-aware recommender systems (TARS) are indeed receiving increasing attention. A wide range of approaches dealing with the time dimension in user modeling and recommendation strategies have been proposed. In the literature, however, reported results and conclusions about how to incorporate and exploit time information within the recommendation processes seem to be contradictory in some cases. Aiming to clarify and address existing discrepancies, in this paper we present a comprehensive survey and analysis of the state of the art on TARS. The analysis show that meaningful divergences appear in the evaluation protocols used—metrics and methodologies. We identify a number of key conditions on offline evaluation of TARS, and based on these conditions, we provide a comprehensive classification of evaluation protocols for TARS. Moreover, we propose a methodological description framework aimed to make the evaluation process fair and reproducible. We also present an empirical study on the impact of different evaluation protocols on measuring relative performances of well-known TARS. The results obtained show that different uses of the above evaluation conditions yield to remarkably distinct performance and relative ranking values of the recommendation approaches. They reveal the need of clearly stating the evaluation conditions used to ensure comparability and reproducibility of reported results. From our analysis and experiments, we finally conclude with methodological issues a robust evaluation of TARS should take into consideration. Furthermore we provide a number of general guidelines to select proper conditions for evaluating particular TARS.
Pub.: 15 Feb '13, Pinned: 28 Jun '17
Abstract: The past decade has witnessed the modern advances of high-throughput technology and rapid growth of research capacity in producing large-scale biological data, both of which were concomitant with an exponential growth of biomedical literature. This wealth of scholarly knowledge is of significant importance for researchers in making scientific discoveries and healthcare professionals in managing health-related matters. However, the acquisition of such information is becoming increasingly difficult due to its large volume and rapid growth. In response, the National Center for Biotechnology Information (NCBI) is continuously making changes to its PubMed Web service for improvement. Meanwhile, different entities have devoted themselves to developing Web tools for helping users quickly and efficiently search and retrieve relevant publications. These practices, together with maturity in the field of text mining, have led to an increase in the number and quality of various Web tools that provide comparable literature search service to PubMed. In this study, we review 28 such tools, highlight their respective innovations, compare them to the PubMed system and one another, and discuss directions for future development. Furthermore, we have built a website dedicated to tracking existing systems and future advances in the field of biomedical literature search. Taken together, our work serves information seekers in choosing tools for their needs and service providers and developers in keeping current in the field. Database URL: http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/search.
Pub.: 20 Jan '11, Pinned: 28 Jun '17