Language NATURAL

As was said in the introduction (Section 2.1), chemical structures are the universal and the most natural language of chemists, but not for computers. Computers woi k with bits packed into words or bytes, and they perceive neither atoms noi bonds. On the other hand, human beings do not cope with bits very well. Instead of thinking in terms of 0 and 1, chemists try to build models of the world of molecules. The models ai e conceptually quite simple 2D plots of molecular sti uctures or projections of 3D structures onto a plane. The problem is how to transfer these models to computers and how to make computers understand them. This communication must somehow be handled by widely understood input and output processes. The chemists way of thinking about structures must be translated into computers internal, machine representation through one or more intermediate steps or representations (sec figure 2-23, The input/output processes defined... 

Plane waves are often considered the most obvious basis set to use for calculations on periodic sy stems, not least because this representation is equivalent to a Fourier series, which itself is the natural language of periodic fimctions. Each orbital wavefimction is expressed as a linear combination of plane waves which differ by reciprocal lattice vectors ... 

Searching of one or more on-line databases is a technique increasingly used ia novelty studies. The use of such databases enables the searcher to combine indexing parameters, including national and international classifications natural language words ia the full text of patents, ia their claims, or ia abstracts suppHed by iaventor and by professional documentation services and indexing systems of various sorts. Because the various patent databases have strengths and weaknesses that complement each other, the use of multiple databases is thus pmdent, and is faciUtated by multifile and cross-file techniques provided by the various on-line hosts. 

Symbolic-to-symbolic transformations are used in various symbohc manipulations, including natural language processing and nrle-based system implementation. See Refs. 54 and 140. 

Commands in Mathematica are given in natural language form such as "Solve" or "Simplify" etc. The format of a command is the word starting with a capital letter and enclosing the argument in square brackets ... 

Feldman R, Regev Y, Hurvitz E, Finkelstein-Landau M. Mining the biomedical literature using semantic analysis and natural language processing techniques Biosilico 2003 1 69-80. 

The next steps consist of the extraction and normalization of terms from the zoned input document. To this end, we apply standard natural language processing techniques and normalize the extracted terms to their canonical form with string manipulations and morphological analysis. The former refers to the treatment of symbols (e.g., dashes), and the latter refers to variations of words due to inflection (e.g., plurals). These steps of information extraction rely on, and make extensive use of, our terminologies and ontologies. 

To a computer scientist, VS is nothing but another text mining, only the bits and bytes stored that contain molecular information adopt a format quite different from natural language and without adequate warning cannot be quickly interpreted. It is not that modem day text does not contain text that is not natural language, but that they are adequately flagged and do not stop the NLP software. For example,... 

Interact with the user through a natural language interface. 

Suppose that we have defined a membership function for the "Low pH" set. Most acid solutions would be, to some degree, members of this fuzzy set. We may want to be able to qualify the description by adding modifiers, such as "very," "slightly," or "extremely" whose use allows us to retain close ties to natural language. The qualifiers that modify the shape of a fuzzy set are known as hedges. We can see the effect of the hedges "very" and "very very" in Figure 8.11. 

The "natural" language of medicinal chemists is chemical structure. When claiming important individual compounds, claim the compounds by chemical structure instead of, for example, International Union of Pure and applied Chemistry (IUPAC) chemical names. This minimizes the possibility that an error in a chemical structure will be missed. 

Variable degree of rigor The extent to which postconditions are written in a formal style or in natural language is optional. We prefer more rather than less rigor because we have found it helps find problems early. 

Whereas code is precise, natural language and ad hoc diagrams are not. How often do groups of analysts or designers discuss requirements around a whiteboard and leave with different interpretations of the problem to be solved or produce reams of documents ridden with latent bugs and inconsistencies Documentation that is concise and accurate is far more likely to be useful. 

A model of object state is used to define a vocabulary of precise terms on which to base an analysis, specification, or design. A well-written document should contain plenty of narrative text in natural language along with illustrative diagrams of all kinds, but the type models are used to make sure that there are no gaps or misunderstandings. More on this in Chapter 5, Effective Documentation. 

Experience with formal notations shows that the precision of the notation tends to flush out these issues at the appropriate stage. Anyone who has spent a day or two in a specification workshop, creating a model for a business or a new system, is aware of how quickly important questions are raised. Many of these issues would never have been thought of had the requirements document been in plain natural language (English or your local favorite). 

Each section contains a natural-language description as well as a more formal one, interspersed as far as possible. The formal fragments are specifically used to clarify parts that might otherwise be unclear. 

Informal text can also be substituted for or used as well as any part of a formal expression. For example, it is useful to write postconditions at least in natural language and then in precise terms if required. 

The contract between you and the client can take the form of a requirements document of the traditional kind written in natural language. But having been through the process of formalization, both sides can be confident that it is achievable and consistent and masks no big unknowns. 

The glossary entries for the actions are the pre- and postconditions (formal or informal) with a general natural-language description. 

As usual, diagrams are not intended to be a substitute for good explanation in your natural language. Instead, they are supposed to complement it and remove the ambiguities. 

Now we can rewrite the description of sell. It should be written both in natural language (more readable) and precise terms (less ambiguous) ... 

Write the spec informally. State in natural language what is required of it. Think in terms of outputs, effects on external objects, and the initial and final internal state. Think also of preconditions — interpreted precisely as the conditions under which this particular postcondition make sense, action return (book) pre — book is lent to someone... 

---