International Business Machines Corporation

Juji Corporation

Michelle Xue Zhou

Fei Wang

Jalal U Mahmud

Huahai Yang

Jeffrey William Nichols

Vikram Aggarwal

Wei Zhang

Nimrod Megiddo

Zhen Wen

Shimei Pan

Yunyao Li

Lin Luo

Eser Kandogan

Barton Allen Smith

Liang Gou

Jie Lu

Shixia Liu

Aditya Pal

Wei Hong Qian

Mengdie Hu

Qiang Zhang

Yang Song

Thomas Schaeck

Nan Cao

Shi Xia Liu

Wei Cai

Eben Merriam Haber

Chang Yan Chi

Keith C Houck

Jilin Chen

Li Tan

Qinying Liao

Yangqiu Song

Huiji Gao

Xiao Xiao Lian

Ying-xin Pan

Siwei Shen

Min Chen

Embodiments relate to user attitude modeling and behavior prediction for a social media network. One aspect includes collecting data relating to previously demonstrated sentiments, opinions, and actions attributed to network users toward a topic. Another aspect includes creating a model from the data, which includes factorizing the actions for behavior inference, factorizing auxiliary content from the network for opinion and sentiment inferences, and applying sentiment and opinion regularization to constrain user preferences on implicit topics to explicit sentiments and explicit opinions. Another aspect includes applying the model to a new user of the network with respect to the topic, and generating a prediction with respect to the user that includes predicting sentiment and opinion as a function of the auxiliary content and feature coefficients learned during a training process, and predicting a future action of the user as a function of the auxiliary content and latent profiles of the topic.

Modeling user attitudes toward a target from social media

The present teaching relates to matching users with items. In one example, information related to one or more users is obtained from one or more sources including an online source. The information is transformed into one or more psychological needs of each user. One or more items are obtained. Each item is associated with abilities to satisfy one or more psychological needs. The one or more users are matched with the one or more items based on the users' psychological needs and the items' abilities to satisfy psychological needs to generate a matching result. The matching result is provided.

Method and system for matching people with choices

The present teaching relates to obtaining information from a user via a bot. In one example, a request is obtained to collect information in connection with a user. A statement is generated to be expressed to the user for facilitating a conversation between the user and the bot based on the request. Information is received in connection with the user and collected during the conversation. The collected information characterizes the user in a plurality of modalities. The collected information is automatically analyzed in the plurality of modalities to obtain an assessment of one or more human traits of the user. A plurality of result summaries are generated based on the assessment. The plurality of result summaries are provided in response to the request.

Method and system for creating interactive inquiry and assessment bots

A pairwise relationship data set with multiple attributes (such as, who, what, when, where, how) and with the what attribute (also called the topic attribute) having a word dimension and a people dimension. The data in the topic dimension of the what attribute relates to topics (including other people) relating to the specific, human, personal relationship between the first person and the second person of the pairwise pair. The what attribute data is derived by processing basis data, which includes correspondence data (that is, the substance of correspondence that the first and second persons participate in, including instant messaging and e-mail exchanges. Pairwise relationship data is displayed to a user in real time during a chat session.

Mapping relationships using electronic communications data

Embodiments relate to relationship modeling and visualization from social media. One aspect includes determining a relationship type of a network-based relationship, between an individual and a network contact of the individual, from at least one social media data source. The relationship type is determined using a relationship model based on relationship types that include operational, personal, and business. Another aspect includes performing timeline based relationship strength segmentation using Group Lasso. The timeline based relationship strength segmentation specifies a past and current strength of the relationship. A further aspect includes predicting a future strength of the relationship using Extended Kalman Filter, and providing, through a visual interface, interactive visual analytics to view and monitor relationship states including the past, current, and future strengths over time.

Modeling and visualizing a dynamic interpersonal relationship from social media

Embodiments of the invention relate to assessing characteristics of a message and a message recipient. A trust model is established to take into account a set of trust antecedents, including characteristics of the messages and properties of the recipients, a set of action motivations, and their contribution to the action-based trust measurement. The assessment(s) is utilized to produce a tangible trust measurement that is employed to gauge the recipient's perception of credibility towards the received message.

Action based trust modeling

Embodiments of the invention relate to user identification in social media for task performance based on past usage and traits. One embodiment includes identifying users of a social media platform for ability of performing a task. Readiness scores are predicted based on past usage of the social media platform. A predicted readiness score includes a value corresponding to whether a user is predicted to be ready for performing the task. Users that are ready to perform the task are determined based on comparing the readiness scores with a threshold. Willingness scores for performing the task by the users determined to be able and ready to perform the task are determined. A willingness score includes a value corresponding to whether a user is estimated to be willing for performing the task. Users that are willing to perform the task are determined based on comparing the willingness scores with a threshold.

Identification for performing tasks in open social media

Embodiments relate to interacting with a collection of user opinion documents associated with a topic. One aspect includes obtaining opinion data for the collection of opinion documents associated with the topic. The opinion data includes one or more features discussed in the opinion documents, one or more key phrases included in each feature, one or more text snippets included in each feature, and at least one sentiment expressed in each text snippet. A visual interface is provided in which a feature summary view of the opinion documents acts a top level of a navigational hierarchy. The visual interface allows user navigation from the top level to a lower level of the navigational hierarchy in order to display more details about a text snippet, as compared to the feature summary view, while continuing to provide the feature summary view.

Crowd-powered self-improving interactive visualanalytics for user-generated opinion data

Improved visual context management techniques are disclosed. By way of example, a technique for updating an existing visual display to incorporate new information comprises obtaining new information requested by a subsequent user query, and dynamically deriving one or more visual transformations for updating at least a portion of the existing display to incorporate the new information, wherein the transformation derivation is modeled as an optimization operation which attempts to balance visual context management constraints and to achieve a desired presentation of intended information.

Optimization-based visual context management

Embodiments relate to optimizing user response rate in social media. One embodiment includes generating a model for modeling a likelihood of social media user response from data collected from social media based on social media user filtering criteria and historical information. Modeling the likelihood includes estimating ability of social media users to perform a particular task based on social media user information including tracking of the social media users, evaluating previous task responses and personality features. Social media users are selected based on estimated ability to perform the particular task for maximizing a response rate.

Optimizing user selection for performing tasks in social networks

A computer processor accesses QA forum activity associated with users of a QA forum, wherein the QA forum activity contains historical questions, historical answers, and profiles, each having elements including QA forum information. The processor accesses historical online social activity corresponding to the users of the QA forum, each historical online social activity having elements including at least social activity information. The processor performs a first similarity calculation and selects an at least one model answerer based on the similarity between elements of a question on the QA forum, and the elements of the QA forum activity associated and corresponding to each of the users of the QA forum. Based on a second similarity calculation, the processor determines a similarity between the QA forum user elements and the at least one model answerer elements, and identifies at least one prospect answerer from the QA forum users.

Smart question routing and answerer growing for online community question-answer services

Visualizing content change of a data collection over time. A topic may be split into multiple linear, non-overlapping sub-topics along a timeline by satisfying a diverse set of semantic, temporal, and visualization constraints simultaneously. For each derived sub-topic, a set of representative keywords may be automatically selected to summarize the main content of the sub-topic.

Temporal topic segmentation and keyword selection for text visualization

Embodiments of the invention relate to optimizing user response rate in social media. One embodiment includes generating a statistical model for modeling a likelihood of social media user response from data collected from social media using a set of available features. The social media users are ranked in an estimation set based on the statistical model. The ranked social media users are placed in a rank ordered list. A sub-interval on the rank ordered list is selected for maximizing the response rate.

Embodiments include a program product and a method for providing responses to questions provided on a social media site. The method includes receiving, via a processor, a user question from a social networking site and decomposing and filtering the user question so that it can be further analyzed. The method also includes generating a list of most closely matched potential responders based on analysis of the user question and sending the most closely matched potential responders the user question. Upon receiving responses back from the most closest matched potential responders, these responses are aggregated by the processor in a final response format.

Growing community of inventors

Saratoga, CA, United States of America

Michelle Xue Zhou