International Business Machines Corporation

Vishal George Palliyathu

Balaji Sankar Kumar

John Kurian

Harpreet Singh

Michael J Iantosca

Rohan S Zunzarrao

Ranjith E Raman

Embodiments relate to an intelligent computer platform to support a chatbot platform. A semantically enriched document is subjected to natural language processing to generate a cache of tokens, and further classify the tokens, including noun and verb tokens. For each verb token, a corresponding intent is generated, and for each noun token a corresponding entity is generated. A relationship between the generated intents and entities is mapped, and a topology representing the mapped relationship is constructed. A primary verb is identified and assigned as a root node in the topology, and an arrangement of entities related to primary verb are identified and assigned as child nodes related to the root node. The constructed topology is consumed to an AI schema for implementation in the chatbot platform to support real-time communication flow.

Natural language artificial intelligence topology mapping for chatbot communication flow

An approach for generating an index in Darwin Information Typing Architecture (DITA) source files. The approach determines one or more similar index entries of a prioritized index. The approach determines a fraternal association exists between at least two of the one or more similar index entries when at least one of the one or more similar index entries occur with one or more similar high priority index entries. Responsive to a determination that a fraternal association exists between at least two of the one or more similar index entries, the approach determines whether an inheritance from a parent exists. The approach determines based, at least in part, on whether a fraternal association exists and whether an inheritance from a parent exists, a Darwin Information Typing Architecture (DITA) index.

Generating index entries in source files

A method and associated systems improve access time of a federated repository that represents a set of individual data repositories as a virtualized aggregated repository. An analyzer module counts the number of entries in each individual repository that are associated with each possible value of a selected concordance parameter. The analyzer stores these counts in a Concordance Frequency Table. When the federated-repository manager receives a data-access request, the analyzer associates the requested data element with a corresponding value of the concordance parameter. The analyzer then uses information stored in the Table to select an optimal sequence in which the federated-repository manager should search the repositories for the requested data. This optimal sequence orders the repositories such that the first repositories to be searched will be those that contain the greatest number of entries associated with the concordance-parameter value of the requested data.

Optimizing data access from a federated repository based on concordance frequency

A computer scans a DITA library to identify DITA topic files. The computer then determines whether the identified DITA file has a concept, task, or reference scheme. Based on determining that the identified DITA topic file has a concept scheme, the computer generates a subject taxonomy. Based on determining that the identified DITA topic file has a task scheme, the computer generates a navigation taxonomy. Based on determining that the identified DITA topic file has a reference scheme, the computer generates a command relational taxonomy. Based on the generated subject, navigation, and command relational taxonomies, the computer generates a DITA file relationship table based on the contextual taxonomy density of the aforementioned taxonomies.

DITA relationship table based on contextual taxonomy density

An approach for generating an index in Darwin Information Typing Architecture (DITA) source files. The approach determines a gerund-noun listing based, at least in part, on one or more files. The approach determines a prioritized index of the gerund-noun listing. The approach determines one or more similar index entries of the prioritized index. The approach determines whether a fraternal association exists between at least two of the one or more similar index entries. Responsive to a determination that a fraternal association exists between at least two of the one or more similar index entries, the approach determines whether an inheritance from a parent exists. The approach determines based, at least in part, on whether a fraternal association exists and whether an inheritance from a parent exists, a DITA index.

Growing community of inventors

Bengaluru, India

Vishal George Palliyathu