Search

Today, there are many channels where consumers can share their ideas about products and services. These opinions are usually in text format due to the nature of the feedback. Sentiment analysis is a topic that has gained importance in recent years, especially in text-based information sources. Aspect-based Sentiment Analysis, which is a more sensitive sentiment analysis technique, is the task of determining the aspect term, aspect category and sentiment class in a sentence. A data set consisting of the comments of restaurant customers presented to the competitors in the Semeval ABSA competition is used in the study. Using Word2vec, Glove, Fastext and Bert methods, the aspect term, aspect category and sentiment class are determined on the data set. The hypothesis is tested whether combining the word vector and the sentence vector can improve classification success for ABSA. In the classification made with four different vector methods, Fasttext method with 0.78 F1 score for the target term, Fasttext with 0.57 F1 score for the target category, and Bert method with 0.76 F1 score for the sentiment class have the most successful results. These results are compared with studies in the literature for different data sets and different languages. As a result, it is determined that Fasttext and Bert representation methods give successful results in sentiment analysis of targetbased Turkish language texts. [ABSTRACT FROM AUTHOR]

Türkiye has experienced the most destructive earthquake disaster in recent years. On February 6, 2023, two earthquakes of more than seven magnitudes occurred in Türkiye, first in Pazarcık and then in Elbistan districts of Kahramanmaraş province. These earthquakes caused the deaths of more than 50 thousand people and billions of dollars in damage. The earthquakes had serious effects on every segment of society, including children. This study, which aimed to examine the opinions of individuals about children in such a big earthquake, examined 3.450 tweets sent in February 2023. While the researchers used Maxqda to analyze tweets regarding hashtags, likes, retweets, and source devices, they used Python for sentiment analysis. The findings highlighted that the words AFAD (Disaster and Emergency Management Presidency), education, and psychology were frequently used as hashtags in tweets about children. In addition, the study observed that tweets with very few likes and retweets did not arouse much interest in people. The findings of the sentiment analysis indicated that the tweets were quite negatively oriented. The study's findings can guide future actions and plans concerning the earthquake and children. [ABSTRACT FROM AUTHOR]

The concept of decision making is the process of choosing one of the available alternatives to achieve the goals and objectives. The problems that are frequently encountered today and that require evaluation of many criteria together are selection problems. Multi-criteria decision-making techniques offer suitable solutions to these selection problems. In this study, CRITIC integrated TOPSIS Method was used to determine which University Hospitals in Türkiye are closest to the ideal. In order to select the most suitable University Hospital in study, the most ideal solution was tried to be determined by considering 13 (thirteen) criteria. The criteria used in the study were first weighted with the CRITIC method, and then these weight values were used when ranking the University Hospitals with the TOPSIS method. Thus, it was aimed to determine the University Hospitals closest to the ideal by weighting the criteria. The intended meaning is for hospitals to be ranked at the top according to the assigned criterion weights. As a result of the study, the most ideal University Hospital was U18 (Erciyes University). Then, the University Hospitals are listed as U17 (Ege University Hospital) and U19 (Hacettepe University Hospital), respectively. University Hospitals that are not close to the ideal are U7 (Bozok University Hospital), U8 (Kafkas University Hospital) and U16 (Duzce University Hospital), respectively. [ABSTRACT FROM AUTHOR]

With the increasing use of technology, people have had tools and environments where they can easily share their opinions on almost any subject. An important part of these ideas is related to investment issues. Twitter stands out in this regard and constitutes an important data source for many studies. In many studies, the ideas received on Twitter are analysed and price predictions are made for many products. On the other hand, dividend retirement and dividend investing have become extremely attractive especially for individuals in the small investor class in recent years, and accordingly, individuals have tended to this investment channel especially for their children and families. Individuals also share on Twitter platform about these investments they have made and exchange ideas with each other. At this point, people who will prefer the investment type mentioned are sometimes unsure about which companies to invest in. The aim of this study is to determine the dividend companies that Twitter users like and comment the most and to create alternative investment portfolios. Thus, it has been focused on the most mentioned, most liked, most commented, and most directed (retweeted) tweet features in the study. Moreover, a total of five portfolios have been created with the companies identified from the data. The return conditions of these portfolios have been given in the context of Modern Portfolio Theory based on the last year returns of the firms. It has been determined that the portfolio created according to the retweet status creates the best return from the portfolios. It is thought that the portfolios presented within the scope of the study and the investment alternatives they contain will benefit consumers who are interested in dividend investing. [ABSTRACT FROM AUTHOR]

Due to constantly changing market conditions and developing technology; The importance of entrepreneurial activity is increasing, especially in developing countries such as Turkey. Likewise, the clustering of entrepreneurial activities according to the places with the same features is becoming more and more important. Because the number of these clusters is less than the provinces, development regions and geographical regions. But because of the common features they have, these clusters; It is an important data source both in promoting / supporting entrepreneurship activities and in identifying / resolving problems that arise in entrepreneurship activities. There are no studies in the literature that cluster all cities according to their entrepreneurial characteristics, with this study, it is aimed to close this gap in the literature. Turkey Ministry of Industry and Technology / Entrepreneur Information System (GBS) located in the “Number of enterprises by cities, net sales and assets” were used as data in the study. Fuzzy k means and SOM ANN (Artificial Neural Networks) methods, one of the clustering methods, have been applied to the published data of 2017 in the GBS database. and Turkey’s entrepreneurship is mapped. According to the fuzzy k means method, 19 cities were collected in 3 different clusters and 62 cities were collected in a single cluster. According to SOM clustering method, 8 cities were collected in 4 different clusters and 73 cities were collected in a single cluster. Finally, remarkable findings have been explained and the similarities / differences between the results of the 2 methods have been interpreted. [ABSTRACT FROM AUTHOR]