Figure 3
The figure shows four circles, and at the top, the title reads “2007 to 2012.” The top three circles are attached to each other. Each circle has nodes in it, and all the nodes inside the circles are marked with an inverted triangle. Among them, the largest circle in the middle is labeled “Computer Science, Interdisciplinary Applications, Computer Science, Information Systems, Language and Linguistics,” and it has five nodes inside it, labeled as follows: “word alignment,” “bleu score,” “translation quality,” “machine translation,” and “sentence pair.” The circle on the left is the second largest and is labeled “Computer Science, Artificial Intelligence,” and it has four nodes labeled as follows: “baseline systems,” “relation classification,” “machine learning methods,” and “machine learning.” All these nine nodes are shown interconnected. The circle on the right is the smallest and is labeled “Engineering, Electrical and Electronic Computer Science, Theory and Methods,” and it has three nodes connected to themselves. The three nodes are labeled as follows: “user interests,” “learning process,” and “reinforcement learning.” At the bottom right, an individual circle labeled “Computer Science, Theory and Methods” is shown with four nodes, marked as circles, and labeled as follows: “object category,” “object detection,” “segmentation masks,” and “pascal voc.”Baidu’s network map of the most frequent topics in scientific publications from 2007 to 2012. Source: Author’s analysis based on data extracted from Web of Science