Presents discussion threads which are shared by any two countries, we can view the network with every single discussion thread exposed as extra nodes. We transform the `country-country’ data into `country-thread-country’ information, and after that break the triad into two `country-thread’ dyads. This can be referred to as a bipartite, or 2-mode network (see refs. 20 and 21 for explanations on operating with 2-mode information). This 2-mode data help us visualise the relationships among nations or discussion threads, and to recognize significant structural properties. Sentiment evaluation The content evaluation is conducted within the MySQL database with custom scripts. Working with the 853 NSC348884 cost messages located inside the network analysis, we perform a sentiment analysis from the messages to determine the opinions of ecigarettes in the neighborhood. To determine if a message is optimistic or negative, we use a very simple bag-of-wordsChu K-H, et al. BMJ Open 2015;5:e007654. doi:10.1136bmjopen-2015-model22 of classifying the terms discovered in each message. The dictionary of words comes in the Multi-Perspective Question Answering (MPQA) Subjectivity Lexicon (http:mpqa.cs.pitt.edu), which identifies 6451 words as constructive or damaging, with an further strong or weak quantifier. In the 853 messages concerning e-cigarettes, you will find over 1.four million words within the text. For every message, we compare just about every word and attempt to match it against the terms inside the MPQA dictionary. If the word isn’t located, we also apply a stemming algorithm to determine if the root word is available. By way of example, afflicted isn’t discovered in the sentiment list, but we are able to stem the word to afflict, that is located within the list. In the event the word, or its stemmed root, is located, we apply a score for the message: Strong, positive = +2 Weak, good = +1 Weak, adverse = -1 Sturdy, negative = -2 Simply because messages can be incredibly different in length, the raw scores are inadequate for comparison. Additionally for the raw scores, we also normalise the scores to control for message size. We conduct several tests to find out how sentiment may well connect with diverse elements within the network. 1st, we examine how sentiment scores for ecigarettes evaluate against topics not associated to ecigarettes applying an independent samples t test. We also use results from the network evaluation to discover PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/21331607 any metrics that may possibly connect country interactions together with the sentiment scores. Outcomes Our final dataset consists of 853 messages posted by members in 37 countries, from July 2005 to April 2012. The amount of posts over time might be observed in figure 1. Network analysis Figure two depicts how nations (represented as nodes, or vertices) are linked to each other. A tie connects two countries if they coparticipate in at the least 1 discussion thread (ie, both postmessages within a single thread). The strength of the tie–depicted visually by the thickness of the line–is higher if the two nations share a presence in several discussion threads. The size in the node represents degree centrality, or the amount of other countries a node is connected to. Within the 2-mode network (figure three), red nodes represent nations and blue nodes represent discussion threads. Every single tie now links a nation with discussion threads that have been posted by members of that nation. Node sizes for every single country (ie, red nodes) are reset so they’re each of the same, but we adjust the discussion threads’ (ie, blue nodes) size primarily based on their betweenness centrality. Betweenness is a network measure 
that indicates how frequentl.