International Business Machines Corporation

Pankaj Dhoolia

Vibha Singhal Sinha

Diptikalyan Saha

Sachindra Joshi

Senthil Kumar Kumarasamy Mani

Mangala Gowri Nanda

Krishna Nandivada Venkata

Debdoot Mukherjee

Dinesh Raghu

Rakesh Mohan

Florian Pinel

Juhnyoung Lee

Saurabh Sinha

Santhosh Babu Kumaran

Monika Gupta

Fenno F Heath, Iii

Aubrey J Rembert

Manisha Dattatraya Bhandar

Anjan Nandy

Manish Satyapal Gupta

Wei Zhao

Biplav Srivastava

Rama Kalyani T Akkiraju

Sugata Ghosal

Rong Liu

Richard Thomas Goodwin

Pietro Mazzoleni

Muhtar B Akbulut

Satish Chandra

Dan O'Connor

Prabir Nandi

Narendran Sachindran

Sweefen Goh

Anup Kumar Ghosh

Senthil K Mani

Anca Andreea Ivan

Asidhara Lahiri

Rohan Raju Padhye

Tilak Mitra

Mangala Gowri, Iii

Maharshi H Desai

Andy James Stoneberg

Dhiraj Madan

Nilay Ghosh

Dipankar Saha

Kathleen Byrnes

Luis Alfonso Lastras-Montano

David Nahamoo

Shao Chun Li

Ajay Gupta

Sampath Dechu

Ram Viswanathan

Robert Leslie Yates

Danish Contractor

Rema Ananthanarayanan

Kamal Bhattacharya

Tian-Jy Chao

Neelamadhav Gantayat

Anil Nigam

Thomas Dean Rosinski

Sreeram Ramakrishnan

Gerhard Sigl

Li Zhu

Shiwa S Fu

Rudrapatna Kallikote Shyamasundar

Prashant Jain

Charbak Chatterjee

Tian J Chao

Gaurav Pandey

David Marston

Vineet Kumar

Venkat Raghavan Ganesh Sekar

Biswaroop Chatterjee

Manas R Kumar Singh

Ramesh Babu Tirukoti

Dennis A Conrad

A plurality of constraints associated with conversational steps implemented by a conversation model is extracted from the conversation model. Using the conversational steps and the constraints, a directed graph is constructed, each node in the directed graph representing a conversational step, each directed edge in the directed graph representing a possible execution path from a first conversational step to a second conversational step. An edge in the graph is populated with flow data denoting a probability associated with the edge. By traversing a portion of the graph, an experience preview is generated, the experience preview demonstrating a user experience of a portion of the conversation model.

Self-contained conversational experience previewing

A system may include a memory and a processor in communication with the memory. The processor may be configured to perform operations. The operations may include generating a content page and selecting an annotation container for the content page. The operations may include configuring the annotation container and inputting content into the annotation container. The operations may include submitting at least one attribute to the annotation container to associate the attribute with the content. The operations may include including the content and the at least one attribute in container data and extracting a model from the container data. The operations may include importing the model into a dialog skill and embedding the dialog skill into a user interface.

Incepting conversational ability in content management systems

In an approach for a cooperative build and content annotation system for conversational design of virtual assistants, a processor formulates a build context based on a build activity of a user. A processor formulates one or more content queries based on the build context. A processor builds a content index by augmenting a text-search index with a neural Information Retrieval (IR) index. A processor searches the content index using the one or more content queries to identify content relevant to the build context. A processor determines at least one recommendation for the user based on heuristic rules applied to the build context and the identified content, wherein each recommendation is a build suggestion or a content annotation suggestion.

Cooperative build and content annotation for conversational design of virtual assistants

A domain-specific document is retrieved. The domain-specific document is related to the performance of a process in the domain and includes a plurality of document spans and a plurality of annotations that annotate the spans. A goal-based framework is generated from the domain-specific document. The generation is based on the plurality of annotations and based on the plurality of document spans. The goal-based framework includes entities and entity relationships related to performance of the process. The goal-based framework is transformed into a platform agnostic dialog model based on the entities and the entity relationships. At least one platform specific dialog model is provided to at least one automated response system based on the platform agnostic dialog model.

Assistant dialog model generation

One embodiment provides a method, including: obtaining a plurality of conversational logs; generating a human agent emulator and a user emulator; providing a workspace for a conversational agent, so that an agent designer generates a conversational specification for the conversational agent, wherein the generating a conversational specification comprises: receiving a selection, by the agent designer, of at least one intent for the conversational agent, wherein the receiving a selection is responsive to the conversational agent workspace providing suggestions for intents; providing at least one suggestion for a dialog node that corresponds to the selected at least one intent; and generating a dialog flow for the conversational agent, wherein the generating comprises iteratively receiving, from the agent designer, selection of at least one aspect and receiving at least one selection of the at least one suggestion for dialog nodes; and providing the conversational agent.

Conversational agent generation

Methods, systems, and computer program products for unsupervised learning of interpretable conversation models from conversation logs are provided herein. A computer-implemented method includes obtaining human-to-human conversation logs; training a deep learning model by (i) learning, in an unsupervised manner, semantic labels for dialog contexts in the multiple human-to-human conversation logs, (ii) mapping the learned semantic labels to query responses across the multiple human-to-human conversation logs, and (iii) inferring one or more entities from the multiple conversation logs based at least in part on the mapping; constructing a human-interpretable conversation model based at least in part on patterns determined via the trained deep learning model; and outputting the human-interpretable conversation model to at least one user.

Unsupervised learning of interpretable conversation models from conversation logs

One embodiment provides a method, including: receiving an input from a first user requesting information; generating a conversation model from a dialog that occurs between the user and a human agent; recording the human agent performing an external action required to respond to the input; mapping steps performed during performance of the external action to conversation slots within the dialog; generating an integrated interpretable conversation model comprising a dialog and action script; receiving, at a conversational agent system, a subsequent input from a second user requesting similar information to the information requested by the first user; and providing, by the conversational agent system, a response to the subsequent input, wherein the providing a response comprises the conversational agent system utilizing the integrated interpretable conversational model to replay (i) the dialog and (ii) the action script using the subsequent input.

External action execution with conversational agent

Provided herein is a system, method, and computer program product for modifying a language conversation model of the language learning system. Modifying the language conversation model includes receiving, using a conversational sub-system, voice inputs. The conversational sub-system converts the voice inputs to voice input data and processes the voice input data. The conversational sub-system detects an error in processing the voice input data and, based at least in part on the error, stores additional data comprising additional voice input data in a memory. The conversational sub-system applies machine learning to the additional data to derive a function that is not enabled within the language conversation model. The conversational sub-system develops an update that enables the language conversation model to implement the function. The update is applied to the language conversation model.

Modifying a language conversation model

Methods, systems, and computer program products for integrated curriculum based math problem generation are provided herein. A computer-implemented method includes generating a first map that represents a plurality of concepts pertaining to at least one subject matter area within a curriculum; generating a second map that represents a plurality of concepts pertaining to math word problems associated with the curriculum; determining portions of the first graph that exhibit a structural match to portions of the second graph; automatically creating a math word problem template by extracting multiple variable portions from one of the one or more math word problems associated with the curriculum; and automatically generating an integrated curriculum math word problem by replacing the multiple extracted variable portions in the template with concepts and relationships derived from the portions of the first graph that exhibit a structural match to the portions of the second graph.

Integrated curriculum based math problem generation

Process management techniques using a representation state transfer architecture include, for example, a method of managing a given process in a data-centric manner in a client/server environment, comprising at least one client and at least one server. In an exemplary method, the server obtains from the client at least one request associated with the given process. In response to the request obtained from the client, the server generates an entity, wherein the generated entity is represented in terms of a data model and a lifecycle model associated therewith, wherein the lifecycle model comprises states and state transitions with each transition caused by one or more activities in the given process, and wherein the given process is represented as a collection of communicating entities. The server sends a response to the at least one request to the client, wherein the response is based at least in part on the generated entity.

Process management using representation state transfer architecture

Methods and arrangements for providing classification for a business process hierarchy. Packaged application objects are received, comprising at least one of the standard application objects and custom objects. At least one of the standard application objects is mapped to at least one node of a business process hierarchy, thereby forming at least one mapping template. At least one of the standard application objects is mapped to at least one of the custom objects to create a first mapping, and at least one of the custom objects is mapped to at least one of the standard application objects to create a second mapping. The at least one mapping template, the first mapping and the second mapping are combined, to create a mapping of at least one of the custom objects to the business process hierarchy. Other variants and embodiments are broadly contemplated herein.

Computer assisted classification of packaged application objects and associated customizations into a business process hierarchy

One embodiment provides a method for generating a process learning graph and a document output from a recorded process for training a cognitive agent, the method comprising: utilizing at least one processor to execute computer code that performs the steps of: obtaining a recording of a process, wherein the recording comprises a demonstration of executing the process; generating, using the recording, the process learning graph, wherein the process learning graph identifies a process flow; generating, using the recording, the document output, wherein the document output comprises process screen transitions and process steps; and providing the process learning graph and the document output to the cognitive agent. Other aspects are described and claimed.

Training a cognitive agent using document output generated from a recorded process

Methods and arrangements for testing a code portion. A contemplated method includes: providing a workspace, at a user interface, for testing a code portion; configuring the workspace to execute the code portion; analyzing the code portion to determine at least one defining aspect of the code portion; based on the determining of at least one defining aspect, determining one or more resources referenced in the code portion; and ensuring availability of the one or more resources referenced in the code portion. Other variants and embodiments are broadly contemplated herein.

Smart programming playgrounds

Methods and arrangements for providing human-language descriptors for computer code. An interface is used to input computer code, and a human-language descriptor is automatically associated with at least one code identifier in the computer program code, wherein the human-language descriptor is obtained to assist a user in determining a meaning of the at least one code identifier. The human-language descriptor is visually displayed along with the at least one code identifier. Other variants and embodiments are broadly contemplated herein.

Understanding computer code with human language assistance

Methods, systems, and articles of manufacture for computing program equivalence based on a hierarchy of program semantics and related canonical representations are provided herein. A method includes defining a program semantic representation for a first program; generating a search query based on said semantic representation for the first program; executing said search query against an index of multiple programs to determine a set of candidate programs among the multiple programs, wherein said index comprises a program semantic representation defined for each of the multiple programs; and identifying one or more programs among the set of candidate programs that match the first program by performing one or more comparisons, across the set of candidate programs, of a canonical representation associated with each program in the identified set of candidate programs to a canonical representation associated with the first program.

Computing program equivalence based on a hierarchy of program semantics and related canonical representations

Growing community of inventors

Uttar Pradesh, India

Pankaj Dhoolia