Method of knowledge extraction through data mining

#The disclosed embodiments relate to data mining methods for determining economically valuable cause effect relationships between objects and properties associated with objects using co-occurrence frequency measurements of semantic terms characterizing observations of properties, effects or behaviors of objects in different environments and using these measurements as object descriptors in calculations determining object similarities. Specifically, these methods may be used to identify new indications of medicines, identify biomarkers associated with disease, identify biomarkers associated with drug effects, quantify disease diagnosis, identify novel drug targets, identify pharmacologic equivalencies of medicines, identify pharmacologic equivalencies between medicines and traditional medicines, identify pharmacologic equivalencies between medicines and Natural products, identify equivalencies between alternate medical procedures, identify risk benefit profiles of medicine combinations, identify targets for antibodies, identify synergies between medicines, identify side effects of medicines, identify risks of experimental medicines, and identify functions of biological networks.

#A core aspect of knowledge is to apply observations of objects in different environments in the identification of relationships between the observed objects or properties associated with the observed objects and to use this understanding in the identification of cause effect relationships. With the amount of electronic information exponentially increasing, there is a growing need for extracting knowledge through data mining.

#The extraction of knowledge through data mining requires identification of relationships among frequently large number of heterogeneous observations regarding attributes, characteristics, color, shape, virtues, merits, capacities, features, diagnostic fingerprints, properties, dependencies, actions, behavior, qualities, nearest neighbor interactions, functional relationships, functions, forces, purpose, effects that are associated with large numbers of objects often having very different units of measurements and scales (molecules, proteins, cells, tissues, organs, organisms, plants, animals, humans, planets, solar systems) and to discern amongst them observations that are meaningful for context recognition (the discernment of close relationships) from observations that are not meaningful for context recognition. [Rajaraman, A.; Ullman, J. D. (2011). “Data Mining”. Mining of Massive Datasets. pp. 1-17.] Likewise the extraction of knowledge trough data mining also requires identification of objects sharing characteristic observation and the translation of this information into knowledge. In this respect, knowledge is a continuum ranging from the implicit end of the continuum such as statistical co-occurrences between objects and observations to more explicit knowledge such as causal relations.

#The present description will be better understood from the following detailed description read in light of the accompanying drawings, wherein:

#FIG. 1 shows a block diagram of a computing apparatus 100 that may be used to practice aspects of the disclosed embodiment;

#FIGS. 2 and 3 show a flow chart in accordance with the disclosed embodiments;

#In at least one aspect of the disclosed embodiments, the techniques disclosed herein may be executed by one or more computers under the control of one or more programs stored on computer readable medium. FIG. 1 shows a block diagram of an exemplary computing apparatus 100 that may be used to practice aspects of the disclosed embodiment. The apparatus 100 may include computer readable program code stored on at least one computer readable medium 102 for carrying out and executing the process steps described herein. The computer readable medium 102 may be a memory of the computing apparatus 100. In alternate aspects, the computer readable program code may be stored in a memory external to, or remote from, the apparatus 100. The memory may include magnetic media, semiconductor media, optical media, or any media which is readable and executable by a computer. Computing apparatus 100 may also include a computer processor 104 for executing the computer readable program code stored on the at least one computer readable medium 102. In at least one aspect, computing apparatus may include one or more input or output devices, generally referred to as a user interface 106 which may operate to allow input to the computing apparatus 100 or to provide output from the computing apparatus 100, respectively.

#FIGS. 2 and 3 illustrate a data mining method 200 according to the disclosed embodiments. The computer processor 104 and the memory 102 including the computer readable program code are configured to cause the computing apparatus 100 to at least perform the data mining method disclosed herein. The data mining method 200 described herein comprises a first step 201 of selecting a finite list of objects and pertinent synonyms for each of the selected objects. In a second step 202, the selection of a finite set of semantic descriptors of observations characterizing and identifying effects, properties and behavior of objects in different environment said object observations selected from the group consisting of properties of matter, actions on composition of matter, activities, functions, behavior, modulation of behavior, qualities of composition of matter, interactions with biological systems, interaction with cellular components, effects on cellular components, pharmacological effects, toxicity, effects on ecological systems, effects on organ systems, effects on living organisms, effects on the environment, chemical structure, composition of matter, utility, economic value, manufacture, effects on disease, effects on social interactions, effects on information processing, societal effects, effects on commercial transactions, effects on information transmission, modulation of activities, modulation of functions, physicochemical properties, clinical utility, and therapeutic indications.

#A third step 203 refers to the determination of the co-occurrence frequency of semantic terms identifying selected objects and the semantic terms identifying selected observations in at least one text containing data base.