Using network science and text analytics to produce surveys in a scientific topic (1506.05690v2)

Abstract: The use of science to understand its own structure is becoming popular, but understanding the organization of knowledge areas is still limited because some patterns are only discoverable with proper computational treatment of large-scale datasets. In this paper, we introduce a network-based methodology combined with text analytics to construct the taxonomy of science fields. The methodology is illustrated with application to two topics: complex networks (CN) and photonic crystals (PC). We built citation networks using data from the Web of Science and used a community detection algorithm for partitioning to obtain science maps of the fields considered. We also created an importance index for text analytics in order to obtain keywords that define the communities. A dendrogram of the relatedness among the subtopics was also obtained. Among the interesting patterns that emerged from the analysis, we highlight the identification of two well-defined communities in PC area, which is consistent with the known existence of two distinct communities of researchers in the area: telecommunication engineers and physicists. With the methodology, it was also possible to assess the interdisciplinary and time evolution of subtopics defined by the keywords. The automatic tools described here are potentially useful not only to provide an overview of scientific areas but also to assist scientists in performing systematic research on a specific topic.

• The paper presents a novel framework that combines network science and text analytics to generate surveys and map scientific fields.
• It applies community detection and keyword extraction to partition complex citation networks in fields like Complex Networks and Photonic Crystals.
• The method distinguishes core interdisciplinary topics from specialized ones, offering scalable tools for organizing scientific literature.

Applying Network Science and Text Analytics for Scientific Surveys

The paper presents a novel framework integrating network science and text analytics aimed at generating surveys and mapping knowledge structures within scientific fields. Two distinct areas—Complex Networks (CN) and Photonic Crystals (PC)—were selected as application domains to demonstrate the utility of the proposed method. Utilizing citation data from the Web of Science, the paper leverages a combination of community detection algorithms and text analysis to generate structured science maps and taxonomies.

The methodology involves constructing citation networks for CN and PC and utilizing community detection to partition these networks into distinct communities or topics. This is followed by the creation of an importance index that extracts salient keywords, which in turn are used to define communities. The topology of the network provides additional insights, allowing the creation of dendrograms that establish hierarchical relationships among subtopics.

Key Findings

1. Complex Networks Analysis: The CN network, comprising 11,063 papers, was partitioned into 22 communities. Central topics included epidemic spreading dynamics, synchronization and coupling, and community structure. Peripheral topics were identified, such as applications of networks in ecological systems, gene regulatory networks, and cooperation dynamics.
2. Photonic Crystals Analysis: The PC network consisted of 20,230 papers, revealing two major communities: telecommunications engineers and physicists. This dual-community structure was underscored by a clear divide, with minimal interaction between them. Topics like photonic crystal fibers and supercontinuum generation emerged prominently in the telecommunications domain, while physicists focused on fundamental photonic crystal science.
3. Community and Keyword Analysis: For both CN and PC, the paper illustrates the utility of accessibility metrics in distinguishing core from peripheral communities, suggesting that central communities are enriched with general, interdisciplinary concepts, while peripheral communities tend to be more specialized.
4. Temporal Trends: The paper tracked the evolution of subtopics within CN and PC, identifying growing interest in areas like network applications for time series in CN, and fiber laser and biosensor technologies in PC.

Implications and Future Directions

The proposed framework addresses the challenge of organizing voluminous scientific literature, thereby assisting researchers in constructing coherent surveys. The paper demonstrates that combining network analysis with text analytics can yield informative science maps that reflect the underlying structure and dynamics of scientific fields.

From a practical standpoint, the methodology provides a scalable tool for visualizing the organization of scientific domains, particularly useful in an era of information deluge. The theoretical implications are significant, suggesting avenues for exploring interdisciplinary connections and potentially fostering collaborations across fields previously regarded as distinct.

Future research could expand on this work by incorporating directional citation networks and examining interdisciplinary links. The framework's adaptability allows for its application across various scientific domains, thus promising a versatile tool for academic and research-oriented investigations. Additionally, integrating more sophisticated NLP techniques could enhance the accuracy of keyword extraction, leading to even more refined taxonomies and science maps.

In summary, this paper introduces an efficient, mixed-method approach for constructing structured surveys and scientific maps, illustrating the potential to considerably enhance the ways in which scientific knowledge is organized and interpreted.