Can the Online Social Networks be used as a Learning Tool? A Case Study in Twitter

Campus SorocabaCampus Sorocaba

Can the Online Social Networks be used Can the Online Social Networks be used as a Learning Tool? as a Learning Tool? A Case Study in A Case Study in TwitterTwitter

Available in:

• Draft in english: http://www.dcomp.sor.ufscar.br/lzaina/papers/CCIS2014_draft.pdf• Final version in english: Communications in Computer and Information Science (Print)

• In portuguese – RBCA Journal: http://www.upf.br/seer/index.php/rbca/article/view/2944

LERIS-Laboratory of Studies in Networks, Innovation and Software

www.leris.sor. ufscar.br

Federal University of São Carlos - Sorocaba, Brazil

Luciana A M Zaina, Tiago Almeida and Guilherme Torres

BrazilBrazil

MotivationMotivation

Communication and collaboration tools such as blogs, wikis and social networks have attracted billions of users, and online social online social networksnetworks has surpassed the emailemail popularity.

Online social networks can encourage students interaction.

Twitter implements the conception of microblogging service.

Short messages (limited to 140 characters)

How to analyze the Twitter messages? Text mining techniques that allows us to identifyidentify and extractextract

patternspatterns from a set of messages.

BrazilBrazil

Paper ObjectivePaper Objective This work empirically examine the employing of TwitterTwitter

to supportsupport the outsideoutside classroom discussions. We have collected datadata through text mining techniques:

find termsterms commonly usedused by teachersteachers and studentsstudents

And comparecompare professor’s messages to the messages sent by students.

The experiment was performed with undergraduate student of a Computer Science course in a Brazilian University.

BrazilBrazil

The ExperimentThe Experiment

Observe the students behavior. the students are active agents.

The course subjects: Web Development, Software Engineering and Entrepreneburship

We collected the messages using the Twitter API Participants:

2 Professors 52 students

The steps…

BrazilBrazil

PreparationPreparation

We talked with the students about the use of Twitter but…

Any kind of recommendation regarding the messages format were made by the professors.

After 4 months, 1,794 messages were collected: 118 posted by professors and 1,676 by the students.

The professors pointed out the most relevant keywords to the teaching objectives and consequently to guide the text mining process.

BrazilBrazil

1st Step – Messages analysis (I)1st Step – Messages analysis (I) From the collected messages we applied the text mining

techniques considering the pointed relevant keywords tokenized the messages and constructed arrays to

track the occorrenceoccorrence and frequencyfrequency of each relevant keyword.

compared two matrices of relevant terms: one from the messages posted by the teachers and other from the students ones.

Almeida and Yamakami

BrazilBrazil

1st Step – Messages analysis (II)1st Step – Messages analysis (II)

By the intersection set: there was a low frequency low frequency of messages forward by

the students from the teachers’ original messages teachers’ original messages or even postingposting messages that contained the relevant termsrelevant terms.

BrazilBrazil

2nd Step – Questionnarie 2nd Step – Questionnarie feedbackfeedback

To understand our findings: a questionnairequestionnaire composed by eight questions was

elaborated and applied to the envolved students. From the 52 students who participated of the

experiment, 38 (73%) filled the questionnaire. Summarizing the answers...

BrazilBrazil

Access of Twitter and messages Access of Twitter and messages sending.sending.

What of the listed terms are you interested in reading about? (The set of terms considered relevant by professors was presented).

the terms differ from the data of the mining process when since the students retweeted only messages containing the word "job".

BrazilBrazil

Access of Twitter and messages Access of Twitter and messages sendingsending

We can observe the passive behavior of the students in Twitter, acting as receivers of messages.

BrazilBrazil

Access and relevance of the Access and relevance of the linkslinks

the messages they received helped to acquire new information about the topic of interest

BrazilBrazil

ConclusionsConclusions

Comparing the results: text mining x questionnaire responses: the students behaved as receivers of information. Although the most of students did not send or forward messages

with the relevant terms, they have agreed that the posts contributed to the acquisition of new and interesting information.

So, we evidenced that despite the messages posted by the professors have motivated the students, they usually not they usually not exchangeexchange the information with their colleagues.

Campus SorocabaCampus Sorocaba

AcknowledgmentsAcknowledgments

Campus SorocabaCampus Sorocaba

Thanks!Thanks!

[email protected]

http://www.dcomp.sor.ufscar.br/lzaina

facebook.com/zaina.lucianafacebook.com/zaina.luciana