Kyo-Joong Oh

User profile extraction from Twitter for personalized news recommendation

Extracting personal profiles from various sources such as purchased items, watched movies, mailing records, etc. is important for recommender systems. For personalized news recommendation, in particular, existing methods mostly utilize information obtainable from the news articles read by the users such as titles, texts, and click-through data. This paper aims to investigate a different method to build personal profiles using the information obtained from Twitter to provide personalized news recommendation service. For a Twitter user, our method utilizes tweets, re-tweets, and hashtags, from which important keywords are extracted to build the personal profile. The usefulness of this method is validated by implementing a prototype news recommendation service and by performing a user study. Using a simple cosine similarity measure, we compare the differences among the user profiles, and also among the recommended news lists, in order to check the discriminative power of the proposed method. The prediction accuracy of news recommendation is measured against a small group of users.

Personalized news recommendation using classified keywords to capture user preference

Recommender systems are becoming an essential part of smart services. When building a news recommender system, we should consider special features different from other recommender systems. Hot news topics are changing every moment, thus it is important to recommend right news at the right time. This paper aims to propose a new model, based on deep neural network, to analyse user preference for news recommender system. The model extracts interest keywords to characterize the user preference from the set of news articles read by that particular user in the past. The model utilizes characterizing features for news recommendation, and applies those to the keyword classification for user preference. For the keyword classification, we use deep neural network for online preference analysis, because adaptive learning is necessary to track changes of hot topics sensitively. The usefulness of our model is validated through experiments. In addition, the accuracy and diversity of the recommendation results is also analysed.

An Intelligent Multi-Sensor Surveillance System for Elderly Care

This paper is an overview of our on-going project that proposes a monitoring system based on various sensors to detect risky situations for the elderly. From the standpoint of the end-user, a video surveillance system equipped with many other sensors can relieve caregivers from the need to keep a vigilant eye on each patient’s movements, while such technology can be effectively used for monitoring elderly people with dementia. Since a camera surveillance system has limits to classify complex human actions, this project aims to design an intelligent healthcare surveillance system, which extends the conventional automated video surveillance system with various additional sensors, to improve the performance of surveillance. The main contributions of our proposed system will be to: (i) minimize human intervention; (ii) detect more complex activities and situations using various sensors and improved sensor fusion techniques; and (iii) design a novel classifier that identifies risky situations with the collected information.

A Chatbot for Psychiatric Counseling in Mental Healthcare Service Based on Emotional Dialogue Analysis and Sentence Generation

There are early studies to attempt users for psychiatric counseling with chatbot. They lead to changes in drinking habit based on intervention approach via chat bot. The application does not consider the users psychiatric status through the conversations, continuous user monitoring, and ethical judgment in the intervention. We contend that more accurate and continuous emotion recognition gives better satisfaction to users who need mental health care. In addition, appropriate clinical psychological response based on ethical responses is as well. We suggest a conversational service for psychiatric counseling that is adapted methodologies to understand counseling contents based on of high-level natural language understanding (NLU), and emotion recognition based on multi-modal approach. The methodologies enable continuous observation of emotional changes sensitively. In addition, the case-based counseling response model that combines ethical judgment model provides a suitable response to clinical psychiatric counseling.

Medical Prognosis Generation from General Blood Test Results Using Knowledge-Based and Machine-Learning-Based Approaches

In this paper, we present two approaches to generate prognosis from general blood test results. The first approach is a knowledge-based approach using ripple-down rules (RDR). The knowledge-based approach with RDR converts knowledge of pathologists into a knowledge base with the minimum intervention of knowledge engineers. The second approach is a machine-learning(ML)-based approach using decision tree, random forest and deep neural network (DNN). The ML-based approach learns patterns of attributes from various cases of general blood test. Our experimental results show that there are indeed some important patterns of the attributes in general blood test results, and they are adequately encoded by the both approaches.

The chatbot feels you - a counseling service using emotional response generation

Early study tries to use chatbot for counseling services. They changed drinking habit of who being consulted by leading them via intervene chatbot. However, the application did not concerned about psychiatric status through continuous conversation with user monitoring. Furthermore, they had no ethical judgment method that about the intervention of the chatbot. We argue that more reasonable and continuous emotion recognition will make better mental healthcare experiment. It will be more proper clinical psychiatric consolation in ethical view as well. This paper suggests a introduce a novel chatbot system for psychiatric counseling service. Our system understands content of conversation based on recent natural language processing (NLP) methods with emotion recognition. It senses emotional flow through the continuous observation of conversation. Also, we generate personalized counseling response from user input, to do this, we use additional constrains to generation model for the proper response generation which can detect conversational context, user emotion and expected reaction.

Paraphrase generation based on lexical knowledge and features for a natural language question answering system

A question answering (QA) system constructs its answers automatically by querying a structured database known as a knowledgebase or an unstructured collection of documents and a set of questions. Paraphrase approaches are widely used to solve paraphrastic problems in natural language QA systems. In machine-learning-based Korean paraphrase, the system requires a large-scale mono/bi-lingual corpus. However, thus far, a well-structured corpus is lack, and it is difficult to get alignment data between Korean and English without noise for entailment. This paper creates paraphrase sentences using synonym knowledge and the various features of full morphemes. The results here demonstrate that the paraphrase quality can be improved by the following features: the morpheme type, the dependencies, and the semantic arguments. The feature of the semantic role labeling (SRL) results can be of assistance when attempting to solve instances of word sense disambiguation (WSD) for lexical replacement in Korean.

Context awareness of social group by topic mining on visiting logs of mobile users in two dimensions

Ubiquitous computing technologies have been developed fast and various decision support systems were proposed. Consecutively, in recently, people are working on an implicit service agent of mobile phone to make people to be provided useful services without paying attentions. However, when people want to buy some products or to go somewhere, people are still going to use Internet to get preference information of group of people. They should put queries and click one or more documents. In this sense, how the implicit services could be perfectly available by only mining on the contextual data obtainable from smart phones? In this paper, we propose a model providing group preference information implicitly through topic-mining on the data obtainable from smart phones. We would introduce developed context awareness model in smart phones and apply a topic model which automatically learns a hierarchy of topics to our server. We would also show our simple experimentation to ensure the effectiveness of topic-mining and of provided implicit group preference information.

Controlled dropout: A different approach to using dropout on deep neural network

Deep neural networks (DNNs), which show outstanding performance in various areas, consume considerable amounts of memory and time during training. Our research led us to propose a controlled dropout technique with the potential of reducing the memory space and training time of DNNs. Dropout is a popular algorithm that solves the overfitting problem of DNNs by randomly dropping units in the training process. The proposed controlled dropout intentionally chooses which units to drop compared to conventional dropout, thereby possibly facilitating a reduction in training time and memory usage. In this paper, we focus on validating whether controlled dropout can replace the traditional dropout technique to enable us to further our research aimed at improving the training speed and memory efficiency. A performance comparison between controlled dropout and traditional dropout is carried out by implementing an image classification experiment on data comprising handwritten digits from the MNIST dataset (Mixed National Institute of Standards and Technology dataset). The experimental results show that the proposed controlled dropout is as effective as traditional dropout. Furthermore, the experimental result implies that controlled dropout is more efficient when an appropriate dropout rate and number of hidden layers are used.

Modeling long-term human activeness using recurrent neural networks for biometric data

BackgroundWith the invention of fitness trackers, it has been possible to continuously monitor a user’s biometric data such as heart rates, number of footsteps taken, and amount of calories burned. This paper names the time series of these three types of biometric data, the user’s “activeness”, and investigates the feasibility in modeling and predicting the long-term activeness of the user.MethodsThe dataset used in this study consisted of several months of biometric time-series data gathered by seven users independently. Four recurrent neural network (RNN) architectures–as well as a deep neural network and a simple regression model–were proposed to investigate the performance on predicting the activeness of the user under various length-related hyper-parameter settings. In addition, the learned model was tested to predict the time period when the user’s activeness falls below a certain threshold.ResultsA preliminary experimental result shows that each type of activeness data exhibited a short-term autocorrelation; and among the three types of data, the consumed calories and the number of footsteps were positively correlated, while the heart rate data showed almost no correlation with neither of them. It is probably due to this characteristic of the dataset that although the RNN models produced the best results on modeling the user’s activeness, the difference was marginal; and other baseline models, especially the linear regression model, performed quite admirably as well. Further experimental results show that it is feasible to predict a user’s future activeness with precision, for example, a trained RNN model could predict–with the precision of 84%–when the user would be less active within the next hour given the latest 15 min of his activeness data.ConclusionsThis paper defines and investigates the notion of a user’s “activeness”, and shows that forecasting the long-term activeness of the user is indeed possible. Such information can be utilized by a health-related application to proactively recommend suitable events or services to the user.