Using Social Computing For Knowledge Translation: Exploiting Social Network And Semantic Content Analyses To Facilitate Online Knowledge Translation Within An Online Social Community Of Medical Practitioners
Notice bibliographique
Résumé
Social computing has led to new approaches for Knowledge Translation (KT) by overcoming the temporal and geographical barriers experienced in face-to-face KT settings. Social computing based discussion forums allow the formulation of communities of practice whereby a group of professionals disseminate their knowledge and experiences through online discussions on specialized topics. In order to successfully build an online community of practice, it is important to improve the connectivity between like-minded community members and between like-topic discussions. In this paper we present a Medical Online Discussion Analysis and Linkages (MODAL) method to identify affinities between members of an online social community by applying: (a) social network analysis to understand their social communication patterns during KT; and (b) semantic content analysis to establish affinities between different discussions and professionals based on their communicated content. Our approach is to establish linkages between users and discussions at the semantic and contextual levels—i.e. we do not just link discussions that share exact medical terms, rather we link practitioners and discussions that share semantically and contextually similar medical terms, thus accounting for vocabulary variations, concept hierarchies and specialized clinical scenarios. MODAL incorporates two novel semantic similarity methods to analyze online discussions using: (i) the Generalized Vector Space Model (GVSM) that leverages semantic and contextual similarity to find similarities between discussion threads and between practitioners; and (ii) an extension of the Balanced Genealogy Model (BGM) so that we are able to address non-leaf mappings, issues of homonymity noted in medical terminologies, and further contextualization of the similarity measures using information content measures. We have implemented a similarity metric that captures the concept of "interest" between users or threads, i.e., a numeric measure of how interested user A is in user B, or how much of the information contained in thread A is related to thread B. MODAL measures the "interest" one professional has in another professional within the online community, and then uses this metric to identify those professionals that are sought by other professionals for expert advice—the content experts. Furthermore, by incorporating the interest measures with SNA, MODAL is able to identify the content experts within the community, and analyze the content of their conversations to determine their areas of expertise. Given the short and unstructured nature of online communications, we use the MeSH medical lexicons and the medical text analysis tools, i.e. Metamap, to map the unstructured narrative of online discussions to formal medical keywords based on the MeSH lexicon. MODAL is tested on two online professional communities of healthcare practitioners: (a) Pediatric Pain Mailing List is a community of 460 clinicians from around the world--over a four year period 2505 messages were shared on 783 different discussion threads; (b) SURGINET is a community of 865 clinicians from around the world that use the forum to discuss general surgical issues-it contains over 17000 messages on 2111 threads by 231 users. MODAL is able to identify content experts and link like-minded practitioners based on the content of their conversations rather than on direct ties between them.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,004 | 0,016 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,000 |
| Méta-épidémiologie (sens large) | 0,001 | 0,001 |
| Bibliométrie | 0,012 | 0,008 |
| Études des sciences et des technologies | 0,002 | 0,002 |
| Communication savante | 0,003 | 0,008 |
| Science ouverte | 0,001 | 0,004 |
| Intégrité de la recherche | 0,001 | 0,001 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,002 | 0,001 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».