Evans, James and Pedro Aceves. 2016. “Machine Translation: Mining Text for Social Theory”. Annual Review of Sociology 42:21-50. DOI: 10.1146/annurev-soc-081715-074206
It is interesting to read about the many applications of NLP, IE, IR and ML in the text. This makes me wonder of how these could be applied to my research project.
I will be doing a research which explores the ways in which Islamic social movement organizations (SMO) have asserted themselves as major players in shaping Indonesian politics and in determining the country's democratic direction.
In the project I will form a narrative on how the SMO's relate to the government / state (opposition / ally / neutral) and political parties (forming alliances / denouncing parties / neutral).
I will identify SMOs and actors related to each SMOs, then I will use an Indonesian English newspaper's archive to determine the SMO's relationship to the government and political parties as mentioned above.
Is this a form of supervised learning because I have identified concepts already at hand?
Is looking at the SMOs relationship with the state and political parties a suitable approach for my research?
Is news archive appropriate for this purpose (especially that there's only one newspaper in English, or should I use newspapers in Indonesian language as well, which then pose minimum-resource language problem for me)?
Should I incorporate the SMOs organizational documents (again, all of them are in Indonesian language) and analyze this along with the news?
It is interesting to read about the many applications of NLP, IE, IR and ML in the text. This makes me wonder of how these could be applied to my research project.
I will be doing a research which explores the ways in which Islamic social movement organizations (SMO) have asserted themselves as major players in shaping Indonesian politics and in determining the country's democratic direction.
In the project I will form a narrative on how the SMO's relate to the government / state (opposition / ally / neutral) and political parties (forming alliances / denouncing parties / neutral).
I will identify SMOs and actors related to each SMOs, then I will use an Indonesian English newspaper's archive to determine the SMO's relationship to the government and political parties as mentioned above.
Is this a form of supervised learning because I have identified concepts already at hand?
Is looking at the SMOs relationship with the state and political parties a suitable approach for my research?
Is news archive appropriate for this purpose (especially that there's only one newspaper in English, or should I use newspapers in Indonesian language as well, which then pose minimum-resource language problem for me)?
Should I incorporate the SMOs organizational documents (again, all of them are in Indonesian language) and analyze this along with the news?