Labeled Latent Dirichlet Allocation Research Articles

Research on Chinese Consumers' Attitudes Analysis of Big-Data Driven Price Discrimination Based on Machine Learning.

From the end of 2018 in China, the Big-data Driven Price Discrimination (BDPD) of online consumption raised public debate on social media. To study the consumers’ attitude about the BDPD, this study constructed a semantic recognition frame to deconstruct the Affection-Behavior-Cognition (ABC) consumer attitude theory using machine learning models inclusive of the Labeled Latent Dirichlet Allocation (LDA), Long Short-Term Memory (LSTM), and Snow Natural Language Processing (NLP), based on social media comments text dataset. Similar to the questionnaires published results, this article verified that 61% of consumers expressed negative sentiment toward BDPD in general. Differently, on a finer scale, this study further measured the negative sentiments that differ significantly among different topics. The measurement results show that the topics “Regular Customers Priced High” (69%) and “Usage Intention” (67%) occupy the top two places of negative sentiment among consumers, and the topic “Precision Marketing” (42%) is at the bottom. Moreover, semantic recognition results that 49% of consumers’ comments involve multiple topics, indicating that consumers have a pretty clear cognition of the complex status of the BDPD. Importantly, this study found some topics that had not been focused on in previous studies, such as more than 8% of consumers calling for government and legal departments to regulate BDPD behavior, which indicates that quite enough consumers are losing confidence in the self-discipline of the platform enterprises. Another interesting result is that consumers who pursue solutions to the BDPD belong to two mutually exclusive groups: government protection and self-protection. The significance of this study is that it reminds the e-commerce platforms to pay attention to the potential harm for consumers’ psychology while bringing additional profits through the BDPD. Otherwise, the negative consumer attitudes may cause damage to brand image, business reputation, and the sustainable development of the platforms themselves. It also provides the government supervision departments an advanced analysis method reference for more effective administration to protect social fairness.

Driving Style Recognition under Connected Circumstance Using a Supervised Hierarchical Bayesian Model

In recent years, the automated driving system has been known to be one of the most popular research topics of artificial intelligence (AI) and intelligent transportation system (ITS). The journey experience on automated vehicles and the intelligent automated driving system could be improved by individualization driving understanding. Although previous studies have proposed methods for driving styles understanding, the individualization driving classification has not been addressed thoroughly. Therefore, in this study, a supervised method is proposed to understand driving behavioral structure and the latent driving styles by incorporating the prior knowledge. Firstly, a novel method is established for driving behavioral encoding and raw driving data mining. Then, the Labeled Latent Dirichlet Allocation (LLDA) is proposed to understand the latent driving styles from individual driving with driving behaviors. Finally, the Safety Pilot Model Deployment (SPMD) data are used to validate the performance of the proposed model. Experimental results show that the proposed model uncovers latent driving styles effectively and shows good agreement to real situations, which provides theoretical guidance on driving behavior recognition for better individual experience on automated driving vehicles.

Supervised topic modeling for predicting molecular substructure from mass spectrometry

Small-molecule metabolites are principal actors in myriad phenomena across biochemistry and serve as an important source of biomarkers and drug candidates. Given a sample of unknown composition, identifying the metabolites present is difficult given the large number of small molecules both known and yet to be discovered. Even for biofluids such as human blood, building reliable ways of identifying biomarkers is challenging. A workhorse method for characterizing individual molecules in such untargeted metabolomics studies is tandem mass spectrometry (MS/MS). MS/MS spectra provide rich information about chemical composition. However, structural characterization from spectra corresponding to unknown molecules remains a bottleneck in metabolomics. Current methods often rely on matching to pre-existing databases in one form or another. Here we develop a preprocessing scheme and supervised topic modeling approach to identify modular groups of spectrum fragments and neutral losses corresponding to chemical substructures using labeled latent Dirichlet allocation (LLDA) to map spectrum features to known chemical structures. These structures appear in new unknown spectra and can be predicted. We find that LLDA is an interpretable and reliable method for structure prediction from MS/MS spectra. Specifically, the LLDA approach has the following advantages: (a) molecular topics are interpretable; (b) A practitioner can select any set of chemical structure labels relevant to their problem; (c ) LLDA performs well and can exceed the performance of other methods in predicting substructures in novel contexts.

Automated worker skill evaluation for improving productivity based on labeled LDA

This paper proposed automated systems for analyzing elemental processes and for evaluating work skills. The systems use labeled latent Dirichlet allocation (L-LDA) to classify worker motions obtained from sensors into four elemental processes. L-LDA automatically learns characteristic motions, so there is no need to define and identify motion features. The proposed system predicts elemental processes with over 86.9% recall in experiments using the assembly process data. Analyst burden is greatly reduced as compared to systems requiring manual analysis of elemental processes from recorded task data. The system evaluates worker skills based on analyzed time series data for elemental processes in four categories, namely, correctness, stability, speed, and rhythm. As a result, the evaluation system clarifies workers’ strong and weak points in tasks performed in experiments, providing new knowledge that would be unobtainable under conventional evaluation methods. Manufacturing efficiency can be improved by allocating workers based on their strengths, and training efficiency will be improved when workers’ weak areas are revealed.

Using a Diabetes Discussion Forum and Wikipedia to Detect the Alignment of Public Interests and the Research Literature

Diabetes is a chronic disease that affects millions of people worldwide. It is therefore unsurprising that there is a high volume of public discussions, resources, and research tackling various aspects of the disease. This study describes a new method for identifying areas of public interest in issues like diabetes and compares them to the topics being discussed in research. We tested our method by using posts from a popular diabetes discussion forum (DiabeticConnect), pages (articles) about diabetes published on Wikipedia, and the titles and abstracts of research articles about diabetes from the Scopus database. Tags assigned to each post in the discussion forum were used along with the post itself to compute a Labeled Latent Dirichlet Allocation (LLDA) model, which was then used to classify the Wikipedia pages and research articles. The resulting classifications were then used to compare the prevalence of the topics found in the discussion forum with that in the other two sources. The results show that the public interest in diabetes is not necessarily addressed by researchers. More importantly, the alignment and misalignment in the changes in relative interest over the various topics are evidence that LLDA modeling can be useful for comparing a public corpus, like a diabetes forum, and an academic one, like research article titles and abstracts. The success of using LLDA to classify research articles based on the tags assigned to posts in a public discussion forum shows that this a promising method for better understanding how the scientific community responds to public interests and needs.

A topic-sensitive trust evaluation approach for users in online communities

In order to facilitate human decision making, trust evaluation has received widespread attention in many fields, especially for online services. Most of the existing methods consider trust in a person as a value which does not vary across different scenarios without any attention to the distinction of domains or communities where trust is derived. However, the notion of context is a significant and indispensable factor for trust evaluation in practice. Due to the lack of the consideration of context, traditional methods cannot resolve the issue that arises when a highly trustworthy person in one domain is likely to dominate the results of trust assessment in others where the person is in fact less authoritative. To solve this problem, in this paper, we develop a general approach to accomplish topic-sensitive trust evaluation by considering the context of trust. We first propose a general framework which presents the well-organized architecture of topic-sensitive trust evaluation in online communities. Then, a user-topic model is proposed to automatically extract topic data from user-generated content based on the Labeled Latent Dirichlet Allocation (LLDA) model. To compare the topic differences between users, we design a topic coverage function for revealing their trust relationships in diverse topics. Moreover, we employ two traditional methods and extend them to accomplish trust prediction for people with multiple domain knowledge. Experiments based on a real-world dataset show that extended topic-sensitive approaches are more adaptive and accurate than those topic-free trust evaluation approaches, especially when the trust application scenario features multiple topics.

Multi-class twitter data categorization and geocoding with a novel computing framework

This study details the progress in transportation data analysis with a novel computing framework in keeping with the continuous evolution of the computing technology. The computing framework combines the Labeled Latent Dirichlet Allocation (L-LDA)-incorporated Support Vector Machine (SVM) classifier with the supporting computing strategy on publicly available Twitter data in determining transportation-related events to provide reliable information to travelers. The analytical approach includes analyzing tweets using text classification and geocoding locations based on string similarity. A case study conducted for the New York City and its surrounding areas demonstrates the feasibility of the analytical approach. Approximately 700,010 tweets are analyzed to extract relevant transportation-related information for one week. The SVM classifier achieves >85% accuracy in identifying transportation-related tweets from structured data. To further categorize the transportation-related tweets into sub-classes: incident, congestion, construction, special events, and other events, three supervised classifiers are used: L-LDA, SVM, and L-LDA incorporated SVM. Findings from this study demonstrate that the analytical framework, which uses the L-LDA incorporated SVM, can classify roadway transportation-related data from Twitter with over 98.3% accuracy, which is significantly higher than the accuracies achieved by standalone L-LDA and SVM.

A Multi-Label Supervised Topic Model Conditioned on Arbitrary Features for Gene Function Prediction

With the continuous accumulation of biological data, more and more machine learning algorithms have been introduced into the field of gene function prediction, which has great significance in decoding the secret of life. Recently, a multi-label supervised topic model named labeled latent Dirichlet allocation (LLDA) has been applied to gene function prediction, and obtained more accurate and explainable predictions than conventional methods. Nonetheless, the LLDA model is only able to construct a bag of amino acid words as a classification feature, and does not support any other features, such as hydrophobicity, which has a profound impact on gene function. To achieve more accurate probabilistic modeling of gene function, we propose a multi-label supervised topic model conditioned on arbitrary features, named Dirichlet multinomial regression LLDA (DMR-LLDA), for introducing multiple types of features into the process of topic modeling. Based on DMR framework, DMR-LLDA applies an exponential a priori construction, previously with weighted features, on the hyper-parameters of gene-topic distribution, so as to reflect the effects of extra features on function probability distribution. In the five-fold cross validation experiment of a yeast datasets, DMR-LLDA outperforms the compared model significantly. All of these experiments demonstrate the effectiveness and potential value of DMR-LLDA for predicting gene function.

Developing Embedded Taxonomy and Mining Patients' Interests From Web-Based Physician Reviews: Mixed-Methods Approach.

BackgroundWeb-based physician reviews are invaluable gold mines that merit further investigation. Although many studies have explored the text information of physician reviews, very few have focused on developing a systematic topic taxonomy embedded in physician reviews. The first step toward mining physician reviews is to determine how the natural structure or dimensions is embedded in reviews. Therefore, it is relevant to develop the topic taxonomy rigorously and systematically.ObjectiveThis study aims to develop a hierarchical topic taxonomy to uncover the latent structure of physician reviews and illustrate its application for mining patients’ interests based on the proposed taxonomy and algorithm.MethodsData comprised 122,716 physician reviews, including reviews of 8501 doctors from a leading physician review website in China (haodf.com), collected between 2007 and 2015. Mixed methods, including a literature review, data-driven-based topic discovery, and human annotation were used to develop the physician review topic taxonomy.ResultsThe identified taxonomy included 3 domains or high-level categories and 9 subtopics or low-level categories. The physician-related domain included the categories of medical ethics, medical competence, communication skills, medical advice, and prescriptions. The patient-related domain included the categories of the patient profile, symptoms, diagnosis, and pathogenesis. The system-related domain included the categories of financing and operation process. The F-measure of the proposed classification algorithm reached 0.816 on average. Symptoms (Cohen d=1.58, Δu=0.216, t=229.75, and P<.001) are more often mentioned by patients with acute diseases, whereas communication skills (Cohen d=−0.29, Δu=−0.038, t=−42.01, and P<.001), financing (Cohen d=−0.68, Δu=−0.098, t=−99.26, and P<.001), and diagnosis and pathogenesis (Cohen d=−0.55, Δu=−0.078, t=−80.09, and P<.001) are more often mentioned by patients with chronic diseases. Patients with mild diseases were more interested in medical ethics (Cohen d=0.25, Δu 0.039, t=8.33, and P<.001), operation process (Cohen d=0.57, Δu 0.060, t=18.75, and P<.001), patient profile (Cohen d=1.19, Δu 0.132, t=39.33, and P<.001), and symptoms (Cohen d=1.91, Δu=0.274, t=62.82, and P<.001). Meanwhile, patients with serious diseases were more interested in medical competence (Cohen d=−0.99, Δu=−0.165, t=−32.58, and P<.001), medical advice and prescription (Cohen d=−0.65, Δu=−0.082, t=−21.45, and P<.001), financing (Cohen d=−0.26, Δu=−0.018, t=−8.45, and P<.001), and diagnosis and pathogenesis (Cohen d=−1.55, Δu=−0.229, t=−50.93, and P<.001).ConclusionsThis mixed-methods approach, integrating literature reviews, data-driven topic discovery, and human annotation, is an effective and rigorous way to develop a physician review topic taxonomy. The proposed algorithm based on Labeled-Latent Dirichlet Allocation can achieve impressive classification results for mining patients’ interests. Furthermore, the mining results reveal marked differences in patients’ interests across different disease types, socioeconomic development levels, and hospital levels.

Supervised topic models with weighted words: multi-label document classification

Supervised topic modeling algorithms have been successfully applied to multi-label document classification tasks. Representative models include labeled latent Dirichlet allocation (L-LDA) and dependency-LDA. However, these models neglect the class frequency information of words (i.e., the number of classes where a word has occurred in the training data), which is significant for classification. To address this, we propose a method, namely the class frequency weight (CF-weight), to weight words by considering the class frequency knowledge. This CF-weight is based on the intuition that a word with higher (lower) class frequency will be less (more) discriminative. In this study, the CF-weight is used to improve L-LDA and dependency-LDA. A number of experiments have been conducted on real-world multi-label datasets. Experimental results demonstrate that CF-weight based algorithms are competitive with the existing supervised topic models.

RFBoost: An improved multi-label boosting algorithm and its application to text categorisation

The AdaBoost.MH boosting algorithm is considered to be one of the most accurate algorithms for multi-label classification. AdaBoost.MH works by iteratively building a committee of weak hypotheses of decision stumps. In each round of AdaBoost.MH learning, all features are examined, but only one feature is used to build a new weak hypothesis. This learning mechanism may entail a high degree of computational time complexity, particularly in the case of a large-scale dataset. This paper describes a way to manage the learning complexity and improve the classification performance of AdaBoost.MH. We propose an improved version of AdaBoost.MH, called RFBoost. The weak learning in RFBoost is based on filtering a small fixed number of ranked features in each boosting round rather than using all features, as AdaBoost.MH does. We propose two methods for ranking the features: One Boosting Round and Labeled Latent Dirichlet Allocation (LLDA), a supervised topic model based on Gibbs sampling. Additionally, we investigate the use of LLDA as a feature selection method for reducing the feature space based on the maximal conditional probabilities of words across labels. Our experimental results on eight well-known benchmarks for multi-label text categorisation show that RFBoost is significantly more efficient and effective than the baseline algorithms. Moreover, the LLDA-based feature ranking yields the best performance for RFBoost.

Content Coding of Psychotherapy Transcripts Using Labeled Topic Models.

Psychotherapy represents a broad class of medical interventions received by millions of patients each year. Unlike most medical treatments, its primary mechanisms are linguistic; i.e., the treatment relies directly on a conversation between a patient and provider. However, the evaluation of patient-provider conversation suffers from critical shortcomings, including intensive labor requirements, coder error, nonstandardized coding systems, and inability to scale up to larger data sets. To overcome these shortcomings, psychotherapy analysis needs a reliable and scalable method for summarizing the content of treatment encounters. We used a publicly available psychotherapy corpus from Alexander Street press comprising a large collection of transcripts of patient-provider conversations to compare coding performance for two machine learning methods. We used the labeled latent Dirichlet allocation (L-LDA) model to learn associations between text and codes, to predict codes in psychotherapy sessions, and to localize specific passages of within-session text representative of a session code. We compared the L-LDA model to a baseline lasso regression model using predictive accuracy and model generalizability (measured by calculating the area under the curve (AUC) from the receiver operating characteristic curve). The L-LDA model outperforms the lasso logistic regression model at predicting session-level codes with average AUC scores of 0.79, and 0.70, respectively. For fine-grained level coding, L-LDA and logistic regression are able to identify specific talk-turns representative of symptom codes. However, model performance for talk-turn identification is not yet as reliable as human coders. We conclude that the L-LDA model has the potential to be an objective, scalable method for accurate automated coding of psychotherapy sessions that perform better than comparable discriminative methods at session-level coding and can also predict fine-grained codes.

Metadata Topic Harmonization and Semantic Search for Linked‐Data‐Driven Geoportals: A Case Study Using ArcGIS Online

AbstractGeoportals provide integrated access to geospatial resources, and enable both authorities and the general public to contribute and share data and services. An essential goal of geoportals is to facilitate the discovery of the available resources. Such a process relies heavily on the quality of metadata. While multiple metadata standards have been established, data contributers may adopt different standards when sharing their data via the same geoportal. This is especially the case for user‐generated content where various terms and topics can be introduced to describe similar datasets. While this heterogeneity provides a wealth of perspectives, it also complicates resource discovery. With the fast development of the Semantic Web technologies, there is a rise of Linked‐Data‐driven portals. Although these novel portals open up new ways to organize metadata and retrieve resources, they lack effective semantic search methods. This article addresses the two challenges discussed above, namely the topic heterogeneity brought by multiple metadata standards and the lack of established semantic search in Linked‐Data‐driven geoportals. To harmonize the metadata topics, we employ a natural language processing method, namely Labeled Latent Dirichlet Allocation (LLDA), and train it using standardized metadata from Data.gov. With respect to semantic search, we construct thematic and geographic matching features from the textual metadata descriptions, and train a regression model via a human participants experiment. We evaluate our methods by examining their performances in addressing the two issues. Finally, we implement a semantics‐enabled and Linked‐Data‐driven prototypical geoportal using a sample dataset from Esri's ArcGIS Online.

Centroid prior topic model for multi-label classification

Supervised topic models such as labeled latent Dirichlet allocation (L-LDA) have attracted increased attention for multi-label classification. However, they lack considerations of the label frequency of the word (i.e., the number of labels containing the word), which is crucial for classification. To address this problem, we investigate the L-LDA model and then propose an extension, namely centroid prior topic model (CTPM). Class-feature-centroid (CFC) suggests a discriminative label-word vector that takes the label frequency of the word into account. CPTM uses this CFC vector as prior for label-word distributions. Extensive experiments on the Yahoo! dataset have been conducted to evaluate our algorithm. The experimental results demonstrate that CPTM outperforms the existing multi-label classification algorithms on AUC, Macro-F1 and Micro-F1.

Supervised labeled latent Dirichlet allocation for document categorization

Recently, supervised topic modeling approaches have received considerable attention. However, the representative labeled latent Dirichlet allocation (L-LDA) method has a tendency to over-focus on t...