"... In PAGE 11: ... This acoustic-only model may be thought of as a rough approximation to a minimum description length approach to finding highly repeated speech patterns which are likely to be words of the language. Results of the evaluation shown in Table1 indicate that the cross-modal algorithm was able to extract a large proportion of English words from this very difficult corpus (M2), many associated with semantically correct visual models (M3). Typical speech segments in the lexicons included names of all six objects in the study, as well as onomatopoetic sounds... ..."

"... In PAGE 60: ... 4.4 Comparison Table12 summarizes the similarities and di erences between the systems just discussed. They have similar application domains, use the same modalities and have the same need for cross-modal references.... ..."

"... In PAGE 4: ... The concepts of resilience, vulnerability and adaptation have developed over time, and have been used in various ways, often unrelated to the study on HDGEC. Relevant papers that did not use the keywords given in Table1 or did not cite the seeds listed in Table 2 were not retrieved. In sum, while we aimed for the best and most complete set of relevant publications, we might have missed important contributions.... ..."

"... In PAGE 15: ... By trying to understand more about how people retrieve information and by emphasizing representation and reasoning using domain knowledge, researchers pursuing a knowledge-based approach hope to build systems that achieve significantly better retrieval effectiveness than those based on statistical techniques. Table1 lists the basic differences between the statistical and knowledge-based approaches (for more see (16) [Croft93]). The statistical approach provides techniques that can deal with very large databases in a variety of domains and languages, whereas the knowledge- based approach promises to provide techniques for retrieving passages and extracting facts more accurately.... ..."

"... In PAGE 4: ... One can think of this preexisting information as prior knowledge about the task at hand, and the question is: how can neural networks effectively use these quot;hints quot; (Abu-Mostafa, 1990)? One answer, the KBANN approach (Towell, Shavlik, amp; Noordewier, 1990; Towell, 1992), creates knowledge-based artificial neural networks by producing neural networks whose topological structure matches the dependency structure of the rules in an approximately-correct quot;domain theory quot; (a collection of inference rules about the current task). Table1 shows the correspondences between a domain theory and a neural network, and Figure 2 alcontains a simple example of the K approach to mapping a domain theory into a neural networks. KBANN has been applied to successfully refining domain theories for real-world problems such as gene finding (Towell et al.... ..."

"... In PAGE 4: ... One can think of this preexisting information as prior knowledge about the task at hand, and the question is: how can neural networks effectively use these quot;hints quot; (Abu-Mostafa, 1990)? One answer, the KBANN approach (Towell, Shavlik, amp; Noordewier, 1990; Towell, 1992), creates knowledge-based artificial neural networks by producing neural networks whose topological structure matches the dependency structure of the rules in an approximately-correct quot;domain theory quot; (a collection of inference rules about the current task). Table1 shows the correspondences between a domain theory and a neural network, and Figure 2 alcontains a simple example of the K approach to mapping a domain theory into a neural networks. KBANN has been applied to successfully refining domain theories for real-world problems such as gene finding (Towell et al.... ..."

"... In PAGE 4: ... One can think of this preexisting information as prior knowledge about the task at hand, and the question is: how can neural networks effectively use these quot;hints quot; (Abu-Mostafa, 1990)? One answer, the KBANN approach (Towell, Shavlik, amp; Noordewier, 1990; Towell, 1992), creates knowledge-based artificial neural networks by producing neural networks whose topological structure matches the dependency structure of the rules in an approximately-correct quot;domain theory quot; (a collection of inference rules about the current task). Table1 shows the correspondences between a domain theory and a neural network, and Figure 2 alcontains a simple example of the K approach to mapping a domain theory into a neural networks. KBANN has been applied to successfully refining domain theories for real-world problems such as gene finding (Towell et al.... ..."

"... In PAGE 4: ... One can think of this preexisting information as prior knowledge about the task at hand, and the question is: how can neural networks effectively use these quot;hints quot; (Abu-Mostafa, 1990)? One answer, the KBANN approach (Towell, Shavlik, amp; Noordewier, 1990; Towell, 1992), creates knowledge-based artificial neural networks by producing neural networks whose topological structure matches the dependency structure of the rules in an approximately-correct quot;domain theory quot; (a collection of inference rules about the current task). Table1 shows the correspondences between a domain theory and a neural network, and Figure 2 alcontains a simple example of the K approach to mapping a domain theory into a neural networks. KBANN has been applied to successfully refining domain theories for real-world problems such as gene finding (Towell et al.... ..."