Digital Media Hotspot Mining Algorithm Implementation with Complex Systems in the Mobile Internet Environment

With the development of the Internet, the amount of information present on the network has grown rapidly, leading to increased difficulty in obtaining effective information. Especially for individuals, enterprises, and institutions with a large amount of information, it is an almost impossible task to integrate and analyze Internet information with great difficulty just by human resources. Internet hot events mining and analysis technology can effectively solve the above problems by alleviating information overload, integrating redundant information, and refining core information. In this paper, we address the above problems and research hot event topic sentence generation techniques in the field of hot event mining and design a hybrid event candidate set construction algorithm based on topic core word mapping and event triad selection. The algorithm uses the PAT-Tree technique to extract high-frequency core words in topic hotspots and maps the high-frequency words into sentences to generate a part of event core sentences. The other part of event core sentences is extracted from the topic hotspots by making event triples as candidate elements, and sentences containing event elements are extracted from the topic hotspots. The sets of event core sentences generated by the two methods are mixed and filtered and sorted to obtain the candidate set, which can be used to build a word graph-based main service channel (MSC) model. In this paper, we also propose an improved word graph-based MSC model and use it for the extraction of event topic sentences. Based on the above research, a hot event analysis system is implemented. The system analyzes the existing topic data and uses the event topic sentence generation algorithm studied in this paper to generate the titles of hot spots, that is, hot events. At the same time, the topics are displayed from different dimensions, and data visualization is completed. The visualization includes the trend change of event hotness, trend change of event sentiment polarity, and distribution of event article sources.

Studi Persistensi Suhu Permukaan Laut Tinggi (>30°C) di Perairan Selat Malaka

Hot Event merupakan fenomena suhu permukaan laut (SPL) tinggi lebih dari 30°C dan memiliki mekanisme khusus dalam pembentukannya, yaitu kecepatan angin lemah dan radiasi matahari tinggi. Hot Event memiliki peran penting dalam menyumbang bahang pada pacific warm pool di Samudra Pasifik bagian barat dan berperan mengatur variasi iklim global. Indonesia sebagai negara kepulauan memiliki potensi besar melemahkan sirkulasi angin dan potensi kejadian Hot Event. Selat Malaka merupakan selat terpanjang di Indonesia dan berpotensi untuk menjadi area kajian Hot Event. Fenomena SPLtinggi (>30°C) dan konstan selama 13 tahun (2003 – 2015) di Selat Malaka merupakan hal unik untuk dikaji. Penelitian ini bertujuan untuk mengidentifikasi dan mengetahui mekanisme terjadinya SPL tinggi (>30°C) dan konstan selama 13 tahun (2013 - 2015) di Selat Malaka. Metode yang digunakan adalah metode kuantitatif. Data yang digunakan adalah data harian SPL, angin, arus permukaan, radiasi matahari selama 13 tahun serta batimetri. Pengolahan data menggunakan bahasa pemograman untuk memvisualisasi SPL tinggi >30°C, angin lemah <2 m/det, arus, radiasi matahari tinggi 200 W/m² serta data batimetri. Variasi SPL paling tinggi dan konstan terjadi pada musim timur (Agustus) dan paling rendah pada musim barat (Februari). Fenomena SPL tinggi dan konstan di wilayah kajian B terhadap kajian A dan C disebabkan lemahnya kecepatan angin <2 m/det di wilayah B dibandingkan wilayah A dan C, serta didukung dengan tingginya radiasi matahari dan batimetri wilayah kajian B yang relatif dangkal, sehingga proses pemanasan massa air lebih cepat dibandingkan wilayah kajian lainnya. Hot Event is a phenomenon of high sea surface temperature (SST) over 30 °C and it has a unique mechanism in its formation by the lower wind speed and high solar radiation. Hot Event has an important role in contributing heat to the pacific warm pool in the western Pacific Ocean and play a role in regulating global climate variations. Indonesia as an archipelagic country has the potential to weaken wind circulation and potential Hot Event. The Malacca Strait is the longest strait in Indonesia and it is potential for Hot Events. The phenomenon of high SST (>30 °C) and constant for 13 years (2003 - 2015) in the Malacca Strait is unique to be studied. The present research aims to identify and determine the mechanism of the occurrence of high SST (>30 °C) and constant for 13 years (2013 - 2015). The method used in this study is quantitative method. The data used are daily data of SST, wind, surface currents, solar radiation for 13 years, and bathymetry. Programming was used to visualize high SST >30 °C, lower speed winds <2 m/s, currents, high solar radiation 200 W/m² also bathymetry data. The highest and constant variation of SST occurs in the east season (August) and the lowest in the west season (February). The phenomenon of high and stable SST in area B on A and C is due to the low wind speed <2 m/s in region B compared to A and C and it is supported by high solar radiation and shallow bathymetry in area B, so that the heating process of water mass is faster than other areas.

An Experimental Study of Spammer Detection on Chinese Microblogs

With the development of Web 2.0, social media such as Twitter and Sina Weibo have become an essential platform for disseminating hot events. Simultaneously, due to the free policy of microblogging services, users can post user-generated content freely on microblogging platforms. Accordingly, more and more hot events on microblogging platforms have been labeled as spammers. Spammers will not only hurt the healthy development of social media but also introduce many economic and social problems. Therefore, the government and enterprises must distinguish whether a hot event on microblogging platforms is a spammer or is a naturally-developing event. In this paper, we focus on the hot event list on Sina Weibo and collect the relevant microblogs of each hot event to study the detecting methods of spammers. Notably, we develop an integral feature set consisting of user profile, user behavior, and user relationships to reflect various factors affecting the detection of spammers. Then, we employ typical machine learning methods to conduct extensive experiments on detecting spammers. We use a real data set crawled from the most prominent Chinese microblogging platform, Sina Weibo, and evaluate the performance of 10 machine learning models with five sampling methods. The results in terms of various metrics show that the Random Forest model and the over-sampling method achieve the best accuracy in detecting spammers and non-spammers.