File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/evalu/98/p98-2225_evalu.xml
Size: 2,153 bytes
Last Modified: 2025-10-06 14:00:36
<?xml version="1.0" standalone="yes"?> <Paper uid="P98-2225"> <Title>Aligning Articles in TV Newscasts and Newspapers</Title> <Section position="5" start_page="1384" end_page="1384" type="evalu"> <SectionTitle> 4 Experimental Results </SectionTitle> <Paragraph position="0"> To evaluate our approach, we aligned articles in the following TV newscasts and newspapers: * NHK evening TV newscast, and * Asahi newspaper (distributed in the Internet). We used 143 articles of the evening TV newscasts in this experiment. As mentioned previously, articles in the evening TV newscasts were aligned with articles in the evening paper of the same day and in the morning paper of the next day. Figure 7 shows the results of the alignment. In this experiment, the threshold was set to 100. We used two measures for evaluating the results: recall and precision. The recall and the precision are 97% and 89%, respectively.</Paragraph> <Paragraph position="1"> One cause of the failures is abbreviation of words.</Paragraph> <Paragraph position="2"> For example, &quot;shinyo-kinko (credit association)&quot; is abbreviated to &quot;shinkin&quot;. In our method, these words lower the reliability scores. To solve this problem, we would like to improve the alignment performance by using dynamic programming matching method for string matching. (Tsunoda 96) has reported that the results of the alignment were improved by using dynamic programming matching method.</Paragraph> <Paragraph position="3"> In this experiment, we did not align the TV news articles of sports, weather, stock prices, and foreign balsu kaimaku (Inter-high school baseball games start)&quot; exchange. It is because the styles of these kinds of TV news articles are fixed and quite different from those of the others. From this, we concluded that we had better align these kinds of TV news articles by the different method from ours. As a result of this, we omitted TV news articles the title text of which had the special underline for these kinds of TV news articles. For example, Figure 8 shows a special underline for a sports news.</Paragraph> </Section> class="xml-element"></Paper>