Zipping out relevant information

represent as sequences of characters. Experimental investigations of physical processes, for instance, typically produce sequences or time series of data. Other systems, such as DNA and protein sequences or human language, are intrinsically represented as strings of characters. Treating information as sequences of characters helps make this information searchable—a necessary first step in navigating the overwhelming mass of data facing us today. Although the abundance of information and its accessibility represents an important cultural advance, it also introduces a new challenge: retrieving relevant information. Imagine entering the largest library in the world and seeking all the relevant documents on your favorite topic. Without an efficient librarian's help, this task would be difficult if not impossibly hopeless. The references you wanted likely would remain buried under tons of irrelevancies. On a more positive note, the growing body of available data provides an ideal test bed for theoretical constructions and models. This opportunity has stimulated considerable interest from researchers in many different communi-ties—physicists, mathematicians, economists , and statisticians, to name a few. In this spirit, we seek to discover the most suitable tools for examining large masses of data and extracting useful information from it. To accomplish the ambitious task of finding the proverbial needle in a haystack, we must first define what useful or relevant information is and where and how it is coded. This is a nontrivial problem because information means different things in different contexts. Moreover, it has no absolute value, depending instead on the specific filters observers impose on their data. Consider a simple coin-toss experiment. A gambler is probably only interested in the toss's outcome (heads or tails), but a physicist might be interested in whether the outcomes reveal anything about the coin's nature (such as whether it is honest or dishonest). We extract information via a two-step process. The syntactic step is where we first identify the structures present in messages without associating any specific meaning to them. It is only in the second (or semantic) step that comprehension of meaning occurs; it is the step in which we connect the syntactic information to previous experience and knowledge. As an example of this two-step process, consider how to identify the language in which a given text is written. In the first step, we scan through the text and identify syntactic structures: articles, verbs, adjectives, and so on. But only someone who knows the language can carry …