Automatic Detection of Small Groups of Persons, Influential Members, Relations and Hierarchy in Written Conversations Using Fuzzy Logic
📝 Abstract
Nowadays a lot of data is collected in online forums. One of the key tasks is to determine the social structure of these online groups, for example the identification of subgroups within a larger group. We will approach the grouping of individual as a classification problem. The classifier will be based on fuzzy logic. The input to the classifier will be linguistic features and degree of relationships (among individuals). The output of the classifiers are the groupings of individuals. We also incorporate a method that ranks the members of the detected subgroup to identify the hierarchies in each subgroup. Data from the HBO television show The Wire is used to analyze the efficacy and usefulness of fuzzy logic based methods as alternative methods to classical statistical methods usually used for these problems. The proposed methodology could detect automatically the most influential members of each organization The Wire with 90% accuracy.
💡 Analysis
Nowadays a lot of data is collected in online forums. One of the key tasks is to determine the social structure of these online groups, for example the identification of subgroups within a larger group. We will approach the grouping of individual as a classification problem. The classifier will be based on fuzzy logic. The input to the classifier will be linguistic features and degree of relationships (among individuals). The output of the classifiers are the groupings of individuals. We also incorporate a method that ranks the members of the detected subgroup to identify the hierarchies in each subgroup. Data from the HBO television show The Wire is used to analyze the efficacy and usefulness of fuzzy logic based methods as alternative methods to classical statistical methods usually used for these problems. The proposed methodology could detect automatically the most influential members of each organization The Wire with 90% accuracy.
📄 Content
Automatic Detection of Small Groups of Persons, Influential Members, Relations
and Hierarchy in Written Conversations Using Fuzzy Logic
French Pope III 1 , Rouzbeh A. Shirvani 1, Mugizi Robert Rwebangira2, Mohamed Chouikha1, Ayo Taylor2,
Andres Alarcon Ramirez1, Amirsina Torfi1, french.pope@bison.howard.edu, Rouzbeh.asghari@gmail.com,
rweba@scs.howard.edu, mchouikha@howard.edu, ayo.taylor@gmail.com, alarcon27@hotmail.com,
amirsina.torfi@ bison.howard.edu
1Electrical and Computer Engineering, Howard University, Washington, D.C. 20059 USA
2Systems and Computer Science, Howard University, Washington, D.C. 20059 USA
Abstract— Nowadays a lot of data is collected in online forums.
One of the key tasks is to determine the social structure of these
online groups, for example the identification of subgroups
within a larger group. We will approach the grouping of
individual as a classification problem. The classifier will be
based on fuzzy logic. The input to the classifier will be linguistic
features and degree of relationships (among individuals). The
output of the classifiers are the groupings of individuals. We also
incorporate a method that ranks the members of the detected
subgroup to identify the hierarchies in each subgroup. Data
from the HBO television show The Wire is used to analyze the
efficacy and usefulness of fuzzy logic based methods as
alternative methods to classical statistical methods usually used
for these problems. The proposed methodology could detect
automatically
the
most
influential
members
of
each
organization The Wire with 90% accuracy.
Keywords- Fuzzy Logic;
Text Conversations; subgroup
Identification; hierarchy
I.
INTRODUCTION
In the last decade there has been increasing use of online
platforms such as opinion forums, chat groups, and social
networks because of broad access to the internet and people’s
communication needs. This new way of communicating has
allowed people with different customs, cultures, and locations
to get together virtually to interact and sometimes cooperate
around common interests. On the other hand, the motivation
of many e-commerce companies for understanding the
behavior of internet users as well as the interest of some
security agencies for detecting security threats has created the
need for analyzing the data generated by online communities.
In addition, because of the massive use of online
communication tools and large amount of information
generated by their users, it is almost impossible to manually
analyze all of the generated information. Therefore, there have
lately been important efforts that seek to automatically
analyze and extract relevant information from written data
corresponding to dialogues among several persons. One of the
active areas of research is to detect associations among the
members of an online community by subgroup identification
in written conversations. The idea of subgroup identification
is to identify members from a community who have similar
ways of thinking or have the same affiliation and may
cooperate each other. Yessenalina et al. [1] proposed a
methodology that classifies the speaker’s side in a corpus of
congressional floor debates, using the speaker’s final vote on
the bill as a labeling for side. This work infers agreement
between speakers based on cases where one speaker mentions
another by name, and a simple algorithm for determining the
polarity of the sentence in which the mention occurs. Gupte
et al. [2] address the problem of segmenting small group
meetings in order to detect different group configurations in
an intelligent environment. They propose an unsupervised
method based on the calculation of the Jeffrey divergence
between histograms of speech activity observations. These
histograms are generated from adjacent windows of variable
size slid from the beginning to the end of a meeting recording.
Elson et al. [3] proposed a method for detecting social
networks from nineteenth-century British novels and serials.
They linked two characters based on whether or not they
conversed.
Tan et al. [4] proposed an algorithm that seeks to detect
groups of people in Twitter with the same affiliation. To do
this, it assumes that connected users are more likely to hold
similar opinions. Finally, the discussants were classified in
groups based on how often they reply to each other. Kunegis
et al. [5] studied user relationships in the Slashdot technology
news site. Slashdot gives users the option of tagging other
users as friends or foes, providing positive and negative
endorsements. Abu et al. [6] identified subgroups in
ideological discussions. To do this, they identified the
discussion participants, comments, and the reply structure of
the thread (i.e. who replies to whom). Then, they used
sentiment analysis to determine the polarity of the comment
(positive or negative) made by a particular participant. Finally,
to identify the subgroup me
This content is AI-processed based on ArXiv data.