"... In PAGE 3: ... Being a smoker and having a first or second degree tear (compared with no tear) were less associated with AI. Throughout the multiple logistic regression model, being older than 35 years, having a third or fourth degree tear, and having a second stage of labour lasting more than one hour maintained their significance ( Table3 ). Having a third or fourth degree tear increased the risk of having AI four-fold.... ..."

"... In PAGE 2: ... Activities such as financial trading, business negotiations, and forecasting (from weather to politics) meet this criteria. The first column of Table1 shows some of the characteristics of these applications from the AI point of view. We are not claiming that these are the only activities of interest, just that they are important considerations for a wide range of interesting problem domains.... In PAGE 2: ...able 1. Characteristics of AI problems and how they are exhibited by poker. We are currently studying the game of poker and are attempting to build a high- performance poker program that is capable of beating the best human players. As shown in the second column of Table1 , poker exhibits all of the activities we are interested in studying on at least some level. Certain aspects of poker have been extensively studied by mathematicians and economists but, surprisingly, very little work has been done by computing scientists.... In PAGE 3: ... We have chosen to study the game of Texas Hold apos;em, the poker variation used in the annual World Series of Poker Championships. It is considered to be the most strategically difficult poker variant that is widely played, and requires all of the complex activities listed in Table1 . For example, the best risk management strategy in the world cannot compensate for a lack of deception, since human opponents are quick to exploit predictable players, no matter how strong they might otherwise be.... In PAGE 4: ... 3. Requirements for a World-Class Poker Player We have identified several key components (modules) that incorporate some of the required activities of a strong poker player and address most of the six characteristics listed in Table1 . However, these components are not independent.... ..."

"... In PAGE 2: ... Activities such as financial trading, business negotiations, and forecasting (from weather to politics) meet this criteria. The first column of Table1 shows some of the characteristics of these applications from the AI point of view. We are not claiming that these are the only activities of interest, just that they are important considerations for a wide range of interesting problem domains.... In PAGE 2: ...able 1. Characteristics of AI problems and how they are exhibited by poker. We are currently studying the game of poker and are attempting to build a high- performance poker program that is capable of beating the best human players. As shown in the second column of Table1 , poker exhibits all of the activities we are interested in studying on at least some level. Certain aspects of poker have been extensively studied by mathematicians and economists but, surprisingly, very little work has been done by computing scientists.... In PAGE 3: ... We have chosen to study the game of Texas Hold apos;em, the poker variation used in the annual World Series of Poker Championships. It is considered to be the most strategically difficult poker variant that is widely played, and requires all of the complex activities listed in Table1 . For example, the best risk management strategy in the world cannot compensate for a lack of deception, since human opponents are quick to exploit predictable players, no matter how strong they might otherwise be.... In PAGE 4: ... 3. Requirements for a World-Class Poker Player We have identified several key components (modules) that incorporate some of the required activities of a strong poker player and address most of the six characteristics listed in Table1 . However, these components are not independent.... ..."

"... In PAGE 17: ...ther non-instantaneous impromptu information resources, e.g., successively refined Web searches. An interesting open question is whether this outlines a temporal distribution dimension of WF. If so, should enactment engines account for non-instantaneous access and utilization of certain artifacts and resources, and how? And how should this be described in the process model? Table4 in Appendix A provides a synopsis for an application WF in the context of AI2TV. 4.... ..."

"... In PAGE 13: ...Bringing Artificial Intelligence into the Real World Builders of expert systems must choose the most appropriate AI techniques for a particular application ( Table3 ). In addition, developers must address organizational and operational aspects of a decision support system (Table 4).... ..."