Results 1 - 10 of 10,029

Table 5: WERs (%) for the topic sensitive bigram model

in A Maximum Entropy Language Model with Topic Sensitive Features
by Jun Wu, Advisors Dr, Sanjeev Khudanpur, Dr. Frederick Jelinek 1998
"... In PAGE 14: ... If some of them are missed, the topic assignment may not be accurate. We are encouraged that the word error rate is also reduced with the reduction of perplexity ( Table5 ). It is reduced by 0.... ..."
Cited by 1

Table 1: ROUGE-1 and ROUGE-2 scores for different com- binations of topic-sensitivity and edge weighting(u) settings.

in Timestamped Graphs: Evolutionary models of text for multidocument summarization
by Ziheng Lin, Min-yen Kan 2007
"... In PAGE 7: ...em performance. We use concept links (Ye et al., 2005) as the similarity function and a MMR reranker to remove redundancy. Table1 shows the results. We observe that both topic-sensitive Pag- eRank and weighted edges perform better than ge- neric PageRank on unweighted timestamped graphs.... ..."
Cited by 1

Table 1: Perplexity results for topic sensitive bigram language model, different history lengths

in Dynamic Nonlocal Language Modeling via
by Hierarchical Topic-Based Adaptation, Radu Florian, David Yarowsky 1999
Cited by 19

Table 2.6: Topics used in Topic-Sensitive PageRank

in 2.1 The World Wide Web...................... 10 2.2 Crawlers and Search Engines................... 11
by Daniel Olmedilla 2003

Table 7: Ranking preferred by majority of users

in ABSTRACT In
by unknown authors
"... In PAGE 7: ... The average precision for the rankings induced by the topic-sensitive PageRank scores is substantially higher than that of the unbiased PageRank scores. Furthermore, as shown in Table7 , for nearly all queries, a majority of the users preferred the rankings in- duced by the topic-sensitive PageRank scores. These results suggest that the e ectiveness of a query-result scoring func- tion can be improved by the use of a topic-sensitive Page- Rank scheme in place of a generic PageRank scheme.... ..."

Table 2: Table shows the geographical distribution of Twitter users. North America, Europe and Asia have the highest adoption of Twitter.

in Why We Twitter: Understanding Microblogging Usage and Communities
by Akshay Java, Tim Finin
"... In PAGE 4: ... For the 76K users in our collec- tion about 39K had specified locations that could be parsed correctly and resolved to their respective latitude and longi- tudinal coordinates (using Yahoo! Geocoding API7 ). Fig- ure 7 and Table2 shows the geographical distribution of Twitter users and the number of users in each continent. Twitter is most popular in US, Europe and Asia (mainly Japan).... ..."

Table 6. Ranking scheme preferred by majority of users.

in Topic-Sensitive PageRank: A Context-Sensitive Ranking Algorithm for Web Search
by Taher H. Haveliwala 2003
"... In PAGE 10: ...igher than that of the unbiased PageRank scores; 0.51 vs. 0.28. Furthermore, as shown in Table6 , for nearly all queries, a majority of the users selected the rankings induced by the topic-sensitive PageRank scores as the better of the two. These results suggest that the effectiveness of a query-result scoring function can be improved by the use of a topic-sensitive PageRank scheme in place of a generic PageRank scheme.... ..."
Cited by 39

Table 6. Ranking scheme preferred by majority of users.

in Topic-sensitive pagerank: A context-sensitive ranking algorithm for web search
by Taher H. Haveliwala 2003
"... In PAGE 10: ...igher than that of the unbiased PageRank scores; 0.51 vs. 0.28. Furthermore, as shown in Table6 , for nearly all queries, a majority of the users selected the rankings induced by the topic-sensitive PageRank scores as the better of the two. These results suggest that the e ectiveness of a query-result scoring function can be improved by the use of a topic-sensitive PageRank scheme in place of a generic PageRank scheme.... ..."
Cited by 39

Table 2: The mean weighted relevance values and their 99.9% confi- dence intervals, given in descending order.

in Assessment of WWW-Based Ranking Systems for Smaller Web Sites
by Ola Ågren
"... In PAGE 4: ...1.2 Weighted The same four groups with overlapping confidence in- tervals can be found in the weighted result set as well as can be seen in Figure 1 and Table2 . The first group consists of HyS2ProT, S2ProT and ProT, the second of Topic-Sensitive PageRank, the third of both versions of HITS Authority and the final group contains (Random- ized) HITS Hub.... ..."

Table 2: Performance comparison and analysis Metric PR TSPR FR

in General Terms Algorithms, Experimentation
by Tony Abou-assaleh, Tapajyoti Das, Weizheng Gao, Yingbo Miao, Zhen Zhen
"... In PAGE 2: ... Table 2(a) shows performance comparisons of the mean average score (MAS) and mean average precision (MAP). Results of the two-tailed t-tests are listed in Table2 (b). Topic-Sensitive PageRank (TSPR) and our FocusedRank (FR) out-perform traditional PageRank (PR) with almost a 5-fold increase in mean average score and a 2-fold increase in mean average precision.... ..."
Results 1 - 10 of 10,029