國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
et al, 2014), representing the overlapping intermedia agenda setting. Through the observations in Figure 6, the similar frequency trends between two media platforms also revealed overlapping intermedia agenda setting, and did not reveal the salient divergence and comparable issue cycles in cross-media contents.
4.2. Semantic Networks Analysis: Facebook Actors and News Media
To answer RQ1, this research applied semantic analysis to analyze the salient concepts of Shenao power plant issue. It would compare the salient issues from communicators on Facebook, traditional media and digital native media. The UCINET 6 calculated the eigenvector centrality as the metric to measure the influence of terms in the semantic network.
Table 6 showed the top-ten words ranked by eigenvector centrality values.
Table 6
Top-Ten Salient Concepts with Greatest Eigenvector Centralities: Facebook Actors and News Media
Rank DPP KMT Environ TNALL DNALL
1 深澳電廠 深澳電廠 深澳電廠 深澳 深澳
245.21 2042.07 1612.66 14705.52 11383.34
2 國民黨 國民黨 深澳 深澳電廠 國民黨
75.68 796.63 295.50 2916.73 1804.83
3 深澳 深澳 燃煤電廠 國民黨 市長
57.96 376.33 240.86 1220.10 993.15
4 停建 民進黨 台灣 燃煤電廠 燃煤電廠
34.27 358.78 165.38 932.21 638.20
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
5 侯友宜 燃煤電廠 政府 市長 民進黨
32.80 278.07 129.49 798.96 547.06
6 市長 台灣 能源 民進黨 深澳電廠
30.72 226.70 120.85 698.43 473.99
7 民進黨 公投 健康 新北 新北
29.97 213.01 102.05 602.16 433.44
8 台灣 新北 環境 政府 強調
27.43 178.14 94.85 535.74 408.08
9 燃煤電廠 健康 環評 政策 政策
23.21 158.55 91.40 503.22 353.07
10 政策 政府 藻礁 台灣 政府
22.47 147.21 79.72 484.53 348.88
Generally, all the communicators focused on Shenao power plant issue, partisan, politics, and environmental concerns. All communicators in RQ1 suggested the “深澳電廠 (Shenao power plant)” or “深澳 (Shenao)” as the most crucial concept, since this was the core word of the whole issue. The “燃煤電廠 (coal-fired power plant)” ranked in the former places as the Shenao power plant was a coal-fired power plant. Besides, except for the environmental groups, the word “國民黨 (Kuo Ming Tang-KMT)” was the secondary concept of all the communicators. Overall, the two political parties referring to “國民黨 (KMT)” and “民進黨 (DPP)” dominated all the communicators except for the environmental groups. In addition, the words “政策 (policy)”, “政府 (Government), or “公投 (referendum)” were ranked in the top-ten words among all communicators, which revealed the communicators’ focus on the development of the policy, how the government made the decisions and the discussion of this issue as a referendum. From the environmental groups, the words related to partisanship was
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
not mentioned on the top-ten salient concepts. Comparing to other communicators, the environmental groups focused more on the environmental concerns, such as “環 境 (environment)”, “能源 (energy source)”, “環評 (environmental impact assessment)”, or “藻礁 (alga reef)”.
Both DPP and KMT actors included the two parties “國民黨 (Kuo Ming Tang)” and “民 進黨 (Democracy Democratic Party)” as crucial mentions. Compared with DPP, KMT focused more on the partisan issue, as the term “民進黨” ranked in the former place, although both political party mentioned “國民黨” in the second place. On the other hand, “停建 (cessation of the construction)” was ranked in the forth place in DPP actors, which depicted the discussion of policy change to call off the Shenao power plant project. Several DPP politicians discussed the “cessation of the construction” of Shenao power plant. Some mentioned the LNG terminal in Taoyuan, which was planned to replace the energy supply after terminating the Shenao power plant; others declared the statements or interests on the Shenao power project due to the cessation. A DPP legislator, Jian-Peng Kao agreed the proposal to cease Shenao power plant as a resut of possible environmental risks. Although the incumbement party agreed with Shenao power plant, some DPP political actors expressed different ideas.
As for the traditional news media and digital native news media, the word “市 長 (mayor)” and “新北 (New Taipei city)” were salient concepts marked in top-ten words. For one thing, as the Shenao power plant was located in New Taipei City, the the statement from local resident, or the discussion of the ecosystem where the Shenao power plant located. For another, how the New Taipei City mayor treated this issue would be another focus. It could be relevant to the news event occurred in the middle of July that the New Taipei City government
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
officially filed an administrative appeal to oppose the Shenao power plant expansion project (Huang. A., 2018), few days after several environmental groups made an protest in front of the Executive cabinet. The incumbent New Taipei City mayor, “朱立倫 Chu, Li-Lun” announced on his Facebook post that “The New Taipei City government filed an administrative appeal today, claiming for withdrawing the Shenao power plant environmental impact assessment”.
The term of “市長 (mayor)” was not only ranked in the news media outlets, but also in the DPP actors. In addition, the term “侯友宜 (Hou You-Ih)” ranked before the “市長 (mayor), which was another salient concept revealed in DPP actors’ posts. It might be the reason that
“侯友宜 (Hou You-Ih)” was the candidate of the next New Taipei City mayor in 2018. Also, he launched 20 posts mentioning Shenao power plant issue during this time who was the top-sixth political actor that generated the most posts related to Shenao issue.
This research utilized Gephi to illustrate the semantic networks, through the Force Atlas and Fruchterman-Reingold algorithm (Cherven, 2013; Bastian et al., 2009), two of the graphing methods set in Gephi to visualize semantic networks. As aforementioned in the methodology section, Force Atlas and Fruchterman-Reingold algorithm (Cherven, 2013; Bastian et al., 2009) were force-directed graphing methods, using the nodes’ attraction and repulsition to relatively place nodes in graphs (Cherven, 2013). Compared to Force Atlas method, Fruchterman-Reingold spreads the nodes out from one another, which the nodes looked more scattered in layouts. Cherven (2013) suggested that the algorithms told the same story based on the same dataset in a visually distinct way. This research tried the three different algorithms to illustrate networks, and selected the Force Atlas and Fruchterman-Reingold to visualize layouts: the former helped to place the strengthened nodes in the middle of the graph, and the latter helped
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
spread the nodes out from one another in layouts. Figure 7 suggested the semantic networks of posts from DPP, KMT and environmental groups actors on Facebook, illustrated by Gephi, in which the node size were weighed based on document frequency of the top-100 words in each network.
Similar to the result analyzed by eigenvector centrality, communicators focused on Shenao power plant issue, partisan, politics, and environmental concerns. The semantic network of KMT focused more on the referendum related issues, such as “公投 (referendum)”,
“連署 (countersign)”, “第二階段 (the second-stage)”, compared with DPP actors. Also, words of “反深澳電廠 (anti-Shenao power plant)”, “反核食 (anti-nuclear-polluted foods)” were strongly related to “公投 (referendum)” and “連署 (countersign)”. The term “第二階段 (the second-stage)” referred to the second-stage countersign of the anti-Shenao power plant referendum, which was initiated by KMT legislators. Meanwhile, the KMT politicians initiated the “反 核 食 (anti-nuclear-polluted foods)” referendum against the food imported from Fukushima, Japan, after its nuclear disaster, and the “反空汙 (anti-air-poluuted)” referendum, to decrease the power generation from thermal power plants year by year. On top of this, the words of “公投 (referendum)” or “連署 (countersign)” were highly discussed by KMT politicians. Besides, the premier “賴清德 (Lai, Ching-Te)”, who announced the calling of Shenao power plaint project, was also highlighted by KMT and DPP actors. Some KMT politicans critized the prime minister Lai Ching-Te’s cancellation of Shenao as the “U-turn”
policy change, which was a political consideration for the 2018 local elections, to win votes when the Shenao’s project was opposed by many local citizens and environmental groups.
Both the KMT and DPP focused more on the political concerns.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 7
Semantic Networks of DPP, KMT, and Environmental Groups Actors
Environmental Group
DPP KMT
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Compared with the environmental groups, the visualization of political actors’ concerns did not emphasize environmental issues such as “空氣 (air)”, “藻礁 (alga reef)”, “能源 (energy source)”, “生態 (ecology)”, “健康 (health)” and so on. These were keywords that most environmental groups highlighted in their posts. For example, they worried that the expansion of Shenao power plant would lead to marine pollutions by destroying nearby alga reefs (“藻礁 (alga reef)”) or air pollutions (“空氣 (air)”) from burning coals for electricity power.
Figure 8 illustrated the semantic networks of news articles by traditional news media and digital native news media. Same as Figure 7, this graph was illustrated by Gephi, using Force Atlas to place the strengthened nodes in the middle of the graph and Fruchterman-Reingold to spread the nodes out from one another in layouts. In addition, the node size was weighed based on document frequency of the top-100 words in each network.
Generally, the salient concepts of these two types of media revealed similar issue focus, as the eigenvector centrality result suggested. Compared to the network of traditional media, digital native media emphasized more on two political actors, “蘇貞昌 (Su, Tseng Chang)” and
“侯友宜 (Hou, You Ih)”. Since they were the nominees of New Taipei City mayor in 2018 Taiwanese local election, as “蘇貞昌 (Su, Tseng Chang)” represented for DPP and “侯友宜 (Hou, You Ih)” represented for KMT, they were highlighted in news media outlets, especially in digital native ones. The current New Taipei City mayor, 侯友宜 (Hou, You Ih), won the election in 2018. It also suggested that digital native news media focused more on the statement of local election nominees from New Taipei City, where the Shenao power plant located. On the other hand, tranditional media focused more on the issue based concerns, such as “環評 (Environmental Impact Assessment)”, “觀塘 (Kwun Tong)”, and “天然氣 (natural gas)”. As
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the salient concepts of “觀塘 (Kwun Tong)” (a dictrict in Taoyuan), and “天然氣 (natural gas)”
referred to the issue of third liquefied natural gas (LNG) terminal in Taoyuan, it suggested that the this alternative plan replacing Shenao power plant was a focus in traditional media brands.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 8
Semantic Networks of Traditional News Media and Digital Native News Media Traditional News Media
Digital Native News Media
‧
4.3. Partisan Slant of News Media: Network Correlation Analysis Results