2 research outputs found
Meta-Algorithms untuk Meningkatkan Kinerja Klasifikasi dalam Keberhasilan Telemarketing Perbankan
No full textBanyak bank dan perusahaan asuransi berinteraksi dengan pelanggan mereka menggunakan jasa telemarketing agar dapat lebih meningkatkan pendapatan bank dan mencapai target. Kompetisi yang ketat di industri perbankan membuat tenaga penjualan bank mengejar target terutama menjelang akhir tahun. Pelaku telemarketing atau biasa disebut telemarketer pada perbankan bertugas menghubungi nasabah potensial yang dilakukan secara berkala melalui sambungan telepon. Namun tidak semua nasabah mendapatkan respons yang positif, sering kali telemarketing menemui antipati dari nasabah sebelum proses komunikasi dilakukan. Pendekatan yang diusulkan untuk menemukan nasabah potensial adalah dengan cara klasifikasi data nasabah menggunakan pembelajaran mesin (machine learning). Penggunaan algoritme pembelajaran yang dilakukan secara tunggal memungkinkan untuk dikembangkan menggunakan meta-algorithms sehingga akan memaksimalkan kinerja klasifikasi dalam memprediksi target pelanggan atau nasabah yang sesuai dengan kebutuhan perusahaan agar tujuan pemasaran tercapai. Berdasarkan hasil penelitian yang diperoleh, tindakan pra-pemrosesan dan penggunaan meta-algorithms seperti bagging memberikan kontribusi yang baik untuk proses pembelajaran model decission tree pada data telemarketing dengan hasil akurasi terbaik 98,7%
Artificial Intelligence and the Law: The Use of Artificial Intelligence as a Tool to Assist Judges in Deciding Polygamy Cases
No full textThis research aims to discuss the relationship between Artificial Intelligence (AI) and law. The emergence of the idea of using AI as a tool to analyse judges\u27 decisions has generated mixed responses. On the one hand, the use of AI can be used as a tool to objectively ensure legal certainty, but on the other hand the use of AI can displace the legal supremacy of judges in court. This attracts the author\u27s attention to examine the use of AI in analysing legal cases and as a consideration for judges in deciding polygamy cases. Polygamy itself is a very complex case in court. Judges\u27 considerations in deciding polygamy cases do not only consider procedural aspects, but also involve substantial aspects related to the cumulative and alternative conditions of polygamy. As a chatbot-based platform, AI certainly has limited access in analysing the legal complexity in polygamy cases. This research focuses on the analysis of AI in analysing polygamy cases both in terms of legal basis and justice. This research method is normative with a conceptual legal approach, data is obtained by netnography using the ChatGPT/OpenAi platform and analysed using the content analysis method. The results showed that there were two aspects that were considered by the judge in the AI version of the polygamy case. First, the juridical aspect which is based on the polygamy provisions in the Compilation of Islamic Law. The second aspect is the social aspect based on gender justice. In addition to presenting the essence of several laws and regulations, AI also provides complex social analyses with a gender justice perspective with simple and straightforward sentences. However, this does not mean that AI can be an absolute and stand-alone consideration in polygamy licence cases. AI is only a complement that makes it easier for judges to analyse cases. This is because the judge\u27s involvement in cognitive and psychological aspects is still needed in interacting with litigants in court
