Remote social network profiling involves mapping social networks to understand the social composition of that community.  Remote profiling involves capturing a sense of a person or organization / entity or robot / scripted agent through publicly available information.  The "remote" part refers to the fact that this is done at a distance and not in a face-to-face way. 

Social media is comprised of social media platforms.  

These include the following types and more: 

A number of the platforms have application programming interfaces (APIs) that may be accessed for the extraction of partial data (rarely N = all).  Those who want access to the big data pipes can often access the data commercially. 

The NCapture Chrome web browser add-on enables access to Twitter / X currently.  Other technologies may be used to acquire more data for analysis here. 

Generally speaking, based on social network analysis, a social network is comprised of a subpopulation of people, who coalesce based on shared interests and likes (homophily).  Individuals are known as "egos," and sub-groups in that social network are "entities."  Egos and entities are usually represented as nodes or vertices.  Relationships between egos and entities may be depicted using links or edges.  An undirected social network graph has links that do not have arrows on either end.  A directed social network graph enables arrows on either end of the links to indicate followership / following... 

Application 1:  Data extraction from a social media platform (with NCapture and platform credentials)

Various free and for-cost software may be used to capture social network data...analyze them...and export data visualizations.  Most require registering one's identity on the particular platforms first, so the platform owners have a sense of who is accessing their data through their API. 

So what sorts of analyses may be done once the data is acquired? 

One approach may be to capture demographic and other information about each "case node" and include those in classification sheets for data queries (such as in the qualitative cross-tabulation analysis). 

Another approach is to use survey responses, writings, speeches, songs, and other textual content production, in order to profile the ego or entity...by manual coding...or autocoded topic modeling...or autocoded sentiment analysis...or all the above. 

Clustering algorithms may be used to understand who is speaking with whom inferentially. 

Application 4:  Capturing the conversational contents of the social network

It is possible to capture the topics that are most often discussed in the social network. 

It is possible to capture the sentiment that moves through a social network, the positive, the negative, and the mixed sentiments...and then to see what the most positive, most negative, and such sentiments are about topically. 

It is possible to select a name (or date, or location, or symbol, etc.) and capture the conversations around that name. 

Cluster analyses may be used to map the interrelationships between the semantic words used in the social network's textual interchanges. 

Application 5:  Changes over time in the social network

It is possible to sample from various time periods in the social network to understand how the social network has changed over time in terms of foci. 

It is possible to follow the thread of an idea over various time periods in the social network (assuming access to the relevant data). 

Application 6:  Spatial / locational data

Many social media datasets include a data column for latitude and longitude of the location of the individual or entity (or 'bot) sharing data or messaging.  Locational data may be placed in a matrix to find out what messages and ideas are being shared from particular locations. 

Locational data may be used from classification sheets for particular data queries in qualitative crosstab analyses to capture regional contrasts in responses. 

There is also "declared" locational data from social media, and these are where various accounts ostensibly are from.  These may be studied similarly. 

Beyond these initial ideas, there are many other potentialities.  This page was created just to hopefully spark ideas in the users of NVivo and social media.