The reasons that drive someone to commit suicide are complex and their study has attracted the attention of scientists in different domains. Analyzing this phenomenon could significantly improve the preventive efforts. In this paper we present a method for sentiment analysis of suicide notes submitted to the i2b2/VA/Cincinnati Shared Task 2011. In this task the sentences of 900 suicide notes were labeled with the possible emotions that they reflect. In order to label the sentence with emotions, we propose a hybrid approach which utilizes both rule based and machine learning techniques. To solve the multi class problem a rule-based engine and an SVM model is used for each category. A set of syntactic and semantic features are selected for each sentence to build the rules and train the classifier. The rules are generated manually based on a set of lexical and emotional clues. We propose a new approach to extract the sentence’s clauses and constitutive grammatical elements and to use them in syntactic and semantic feature generation. The method utilizes a novel method to measure the polarity of the sentence based on the extracted grammatical elements, reaching precision of 41.79 with recall of 55.03 for an f-measure of 47.50. The overall mean f-measure of all submissions was 48.75% with a standard deviation of 7%.
NLP; sentiment analysis; emotion classification; polarity measurement; machine learning
We present a system to automatically identify emotion-carrying sentences in suicide notes and to detect the specific fine-grained emotion conveyed. With this system, we competed in Track 2 of the 2011 Medical NLP Challenge,14 where the task was to distinguish between fifteen emotion labels, from guilt, sorrow, and hopelessness to hopefulness and happiness.
Since a sentence can be annotated with multiple emotions, we designed a thresholding approach that enables assigning multiple labels to a single instance. We rely on the probability estimates returned by an SVM classifier and experimentally set thresholds on these probabilities. Emotion labels are assigned only if their probability exceeds a certain threshold and if the probability of the sentence being emotion-free is low enough. We show the advantages of this thresholding approach by comparing it to a naïve system that assigns only the most probable label to each test sentence, and to a system trained on emotion-carrying sentences only.
emotion detection; multi-label classification; thresholds; probability estimates
Biological named entity recognition, the identification of biological terms in text, is essential for biomedical information extraction. Machine learning-based approaches have been widely applied in this area. However, the recognition performance of current approaches could still be improved. Our novel approach is to combine support vector machines (SVMs) and conditional random fields (CRFs), which can complement and facilitate each other. During the hybrid process, we use SVM to separate biological terms from non-biological terms, before we use CRFs to determine the types of biological terms, which makes full use of the power of SVM as a binary-class classifier and the data-labeling capacity of CRFs. We then merge the results of SVM and CRFs. To remove any inconsistencies that might result from the merging, we develop a useful algorithm and apply two rules. To ensure biological terms with a maximum length are identified, we propose a maximal bidirectional squeezing approach that finds the longest term. We also add a positive gain to rare events to reinforce their probability and avoid bias. Our approach will also gradually extend the context so more contextual information can be included. We examined the performance of four approaches with GENIA corpus and JNLPBA04 data. The combination of SVM and CRFs improved performance. The macro-precision, macro-recall, and macro-F1 of the SVM-CRFs hybrid approach surpassed conventional SVM and CRFs. After applying the new algorithms, the macro-F1 reached 91.67% with the GENIA corpus and 84.04% with the JNLPBA04 data.
To create a sentiment classification system for the Fifth i2b2/VA Challenge Track 2, which can identify thirteen subjective categories and two objective categories.
We developed a hybrid system using Support Vector Machine (SVM) classifiers with augmented training data from the Internet. Our system consists of three types of classification-based systems: the first system uses spanning n-gram features for subjective categories, the second one uses bag-of-n-gram features for objective categories, and the third one uses pattern matching for infrequent or subtle emotion categories. The spanning n-gram features are selected by a feature selection algorithm that leverages emotional corpus from weblogs. Special normalization of objective sentences is generalized with shallow parsing and external web knowledge. We utilize three sources of web data: the weblog of LiveJournal which helps to improve the feature selection, the eBay List which assists in special normalization of information and instructions categories, and the suicide project web which provides unlabeled data with similar properties as suicide notes.
The performance is evaluated by the overall micro-averaged precision, recall and F-measure.
Our system achieved an overall micro-averaged F-measure of 0.59. Happiness_peacefulness had the highest F-measure of 0.81. We were ranked as the second best out of 26 competing teams.
Our results indicated that classifying fine-grained sentiments at sentence level is a non-trivial task. It is effective to divide categories into different groups according to their semantic properties. In addition, our system performance benefits from external knowledge extracted from publically available web data of other purposes; performance can be further enhanced when more training data is available.
sentiment analysis; suicide note; spanning n-gram; web data; supervised approach
Predicting a protein's structural or functional class from its amino acid sequence or structure is a fundamental problem in computational biology. Recently, there has been considerable interest in using discriminative learning algorithms, in particular support vector machines (SVMs), for classification of proteins. However, because sufficiently many positive examples are required to train such classifiers, all SVM-based methods are hampered by limited coverage.
In this study, we develop a hybrid machine learning approach for classifying proteins, and we apply the method to the problem of assigning proteins to structural categories based on their sequences or their 3D structures. The method combines a full-coverage but lower accuracy nearest neighbor method with higher accuracy but reduced coverage multiclass SVMs to produce a full coverage classifier with overall improved accuracy. The hybrid approach is based on the simple idea of "punting" from one method to another using a learned threshold.
In cross-validated experiments on the SCOP hierarchy, the hybrid methods consistently outperform the individual component methods at all levels of coverage.
Code and data sets are available at
This paper describes the sentiment classification system developed by the Mayo Clinic team for the 2011 I2B2/VA/Cincinnati Natural Language Processing (NLP) Challenge. The sentiment classification task is to assign any pertinent emotion to each sentence in suicide notes. We have implemented three systems that have been trained on suicide notes provided by the I2B2 challenge organizer—a machine learning system, a rule-based system, and a system consisting of a combination of both. Our machine learning system was trained on re-annotated data in which apparently inconsistent emotion assignment was adjusted. Then, the machine learning methods by RIPPER and multinomial Naïve Bayes classifiers, manual pattern matching rules, and the combination of the two systems were tested to determine the emotions within sentences. The combination of the machine learning and rule-based system performed best and produced a micro-average F-score of 0.5640.
sentiment classification; suicidal emotion; natural language processing; machine learning
This paper describes the National Research Council of Canada’s submission to the 2011 i2b2 NLP challenge on the detection of emotions in suicide notes. In this task, each sentence of a suicide note is annotated with zero or more emotions, making it a multi-label sentence classification task. We employ two distinct large-margin models capable of handling multiple labels. The first uses one classifier per emotion, and is built to simplify label balance issues and to allow extremely fast development. This approach is very effective, scoring an F-measure of 55.22 and placing fourth in the competition, making it the best system that does not use web-derived statistics or re-annotated training data. Second, we present a latent sequence model, which learns to segment the sentence into a number of emotion regions. This model is intended to gracefully handle sentences that convey multiple thoughts and emotions. Preliminary work with the latent sequence model shows promise, resulting in comparable performance using fewer features.
natural language processing; text analysis; emotion classification; suicide notes; support vector machines; latent variable modeling
The identification of catalytic residues is an essential step in functional characterization of enzymes. We present a purely structural approach to this problem, which is motivated by the difficulty of evolution-based methods to annotate structural genomics targets that have few or no homologs in the databases. Our approach combines a state-of-the-art support vector machine (SVM) classifier with novel structural features that augment structural clues by spatial averaging and Z-scoring. Special attention is paid to the class imbalance problem that stems from the overwhelming number of non-catalytic residues in enzymes compared to catalytic residues. This problem is tackled by: 1) optimizing the classifier to maximize a performance criterion that considers both type I and type II errors in the classification of catalytic and non-catalytic residues; 2) under-sampling non-catalytic residues before SVM training; and 3) during SVM training, penalizing errors in learning catalytic residues more than errors in learning non-catalytic residues. Tested on four enzyme datasets – one specifically designed by us to mimic the structural genomics scenario and three previously-evaluated datasets – our structure-based classifier is never inferior to similar structure-based classifiers and comparable to classifiers that use both structural and evolutionary features. In addition to evaluation of the performance of catalytic residue identification, we also present detailed case studies on three proteins. This analysis suggests that many false positive predictions may correspond to binding sites and other functional residues. A web server that implements the method, our own-designed database, and the source code of the programs are publicly available at http://www.cs.bgu.ac.il/~meshi/functionPrediction.
Catalytic Residues; Functional Annotation; Support Vector Machine (SVM); Energy Terms; Spatial Averaging; Feature Selection; Class Imbalance
Support Vector Machines (SVMs) provide a powerful method for classification (supervised learning). Use of SVMs for clustering (unsupervised learning) is now being considered in a number of different ways.
An SVM-based clustering algorithm is introduced that clusters data with no a priori knowledge of input classes. The algorithm initializes by first running a binary SVM classifier against a data set with each vector in the set randomly labelled, this is repeated until an initial convergence occurs. Once this initialization step is complete, the SVM confidence parameters for classification on each of the training instances can be accessed. The lowest confidence data (e.g., the worst of the mislabelled data) then has its' labels switched to the other class label. The SVM is then re-run on the data set (with partly re-labelled data) and is guaranteed to converge in this situation since it converged previously, and now it has fewer data points to carry with mislabelling penalties. This approach appears to limit exposure to the local minima traps that can occur with other approaches. Thus, the algorithm then improves on its weakly convergent result by SVM re-training after each re-labeling on the worst of the misclassified vectors – i.e., those feature vectors with confidence factor values beyond some threshold. The repetition of the above process improves the accuracy, here a measure of separability, until there are no misclassifications. Variations on this type of clustering approach are shown.
Non-parametric SVM-based clustering methods may allow for much improved performance over parametric approaches, particularly if they can be designed to inherit the strengths of their supervised SVM counterparts.
The authors present a system developed for the 2011 i2b2 Challenge on Sentiment Classification, whose aim was to automatically classify sentences in suicide notes using a scheme of 15 topics, mostly emotions. The system combines machine learning with a rule-based methodology. The features used to represent a problem were based on lexico–semantic properties of individual words in addition to regular expressions used to represent patterns of word usage across different topics. A naïve Bayes classifier was trained using the features extracted from the training data consisting of 600 manually annotated suicide notes. Classification was then performed using the naïve Bayes classifier as well as a set of pattern–matching rules. The classification performance was evaluated against a manually prepared gold standard consisting of 300 suicide notes, in which 1,091 out of a total of 2,037 sentences were associated with a total of 1,272 annotations. The competing systems were ranked using the micro-averaged F-measure as the primary evaluation metric. Our system achieved the F-measure of 53% (with 55% precision and 52% recall), which was significantly better than the average performance of 48.75% achieved by the 26 participating teams.
natural language processing; sentiment analysis; topic classification; naïve Bayes classifier
Predicting a protein's structural class from its amino acid sequence is a fundamental problem in computational biology. Much recent work has focused on developing new representations for protein sequences, called string kernels, for use with support vector machine (SVM) classifiers. However, while some of these approaches exhibit state-of-the-art performance at the binary protein classification problem, i.e. discriminating between a particular protein class and all other classes, few of these studies have addressed the real problem of multi-class superfamily or fold recognition. Moreover, there are only limited software tools and systems for SVM-based protein classification available to the bioinformatics community.
We present a new multi-class SVM-based protein fold and superfamily recognition system and web server called SVM-Fold, which can be found at . Our system uses an efficient implementation of a state-of-the-art string kernel for sequence profiles, called the profile kernel, where the underlying feature representation is a histogram of inexact matching k-mer frequencies. We also employ a novel machine learning approach to solve the difficult multi-class problem of classifying a sequence of amino acids into one of many known protein structural classes. Binary one-vs-the-rest SVM classifiers that are trained to recognize individual structural classes yield prediction scores that are not comparable, so that standard "one-vs-all" classification fails to perform well. Moreover, SVMs for classes at different levels of the protein structural hierarchy may make useful predictions, but one-vs-all does not try to combine these multiple predictions. To deal with these problems, our method learns relative weights between one-vs-the-rest classifiers and encodes information about the protein structural hierarchy for multi-class prediction. In large-scale benchmark results based on the SCOP database, our code weighting approach significantly improves on the standard one-vs-all method for both the superfamily and fold prediction in the remote homology setting and on the fold recognition problem. Moreover, our code weight learning algorithm strongly outperforms nearest-neighbor methods based on PSI-BLAST in terms of prediction accuracy on every structure classification problem we consider.
By combining state-of-the-art SVM kernel methods with a novel multi-class algorithm, the SVM-Fold system delivers efficient and accurate protein fold and superfamily recognition.
An ensemble of supervised maximum entropy classifiers can accurately detect and identify sentiments expressed in suicide notes. Using lexical and syntactic features extracted from a training set of externally annotated suicide notes, we trained separate classifiers for each of fifteen pre-specified emotions. This formed part of the 2011 i2b2 NLP Shared Task, Track 2. The precision and recall of these classifiers related strongly with the number of occurrences of each emotion in the training data. Evaluating on previously unseen test data, our best system achieved an F1 score of 0.534.
natural language processing; text analysis; emotion classification; suicide notes
This paper describes a system for automatic emotion classification, developed for the 2011 i2b2 Natural Language Processing Challenge, Track 2. The objective of the shared task was to label suicide notes with 15 relevant emotions on the sentence level. Our system uses 15 SVM models (one for each emotion) using the combination of features that was found to perform best on a given emotion. Features included lemmas and trigram bag of words, and information from semantic resources such as WordNet, SentiWordNet and subjectivity clues. The best-performing system labeled 7 of the 15 emotions and achieved an F-score of 53.31% on the test data.
emotion classification; topic classification; suicide; suicide notes; machine learning
We describe the Open University team’s submission to the 2011 i2b2/VA/Cincinnati Medical Natural Language Processing Challenge, Track 2 Shared Task for sentiment analysis in suicide notes. This Shared Task focused on the development of automatic systems that identify, at the sentence level, affective text of 15 specific emotions from suicide notes. We propose a hybrid model that incorporates a number of natural language processing techniques, including lexicon-based keyword spotting, CRF-based emotion cue identification, and machine learning-based emotion classification. The results generated by different techniques are integrated using different vote-based merging strategies. The automated system performed well against the manually-annotated gold standard, and achieved encouraging results with a micro-averaged F-measure score of 61.39% in textual emotion recognition, which was ranked 1st place out of 24 participant teams in this challenge. The results demonstrate that effective emotion recognition by an automated system is possible when a large annotated corpus is available.
emotion recognition; keyword-based model; machine-learning-based model; hybrid model; result integration
Extraction of clinical information such as medications or problems from clinical text is an important task of clinical natural language processing (NLP). Rule-based methods are often used in clinical NLP systems because they are easy to adapt and customize. Recently, supervised machine learning methods have proven to be effective in clinical NLP as well. However, combining different classifiers to further improve the performance of clinical entity recognition systems has not been investigated extensively. Combining classifiers into an ensemble classifier presents both challenges and opportunities to improve performance in such NLP tasks.
We investigated ensemble classifiers that used different voting strategies to combine outputs from three individual classifiers: a rule-based system, a support vector machine (SVM) based system, and a conditional random field (CRF) based system. Three voting methods were proposed and evaluated using the annotated data sets from the 2009 i2b2 NLP challenge: simple majority, local SVM-based voting, and local CRF-based voting.
Evaluation on 268 manually annotated discharge summaries from the i2b2 challenge showed that the local CRF-based voting method achieved the best F-score of 90.84% (94.11% Precision, 87.81% Recall) for 10-fold cross-validation. We then compared our systems with the first-ranked system in the challenge by using the same training and test sets. Our system based on majority voting achieved a better F-score of 89.65% (93.91% Precision, 85.76% Recall) than the previously reported F-score of 89.19% (93.78% Precision, 85.03% Recall) by the first-ranked system in the challenge.
Our experimental results using the 2009 i2b2 challenge datasets showed that ensemble classifiers that combine individual classifiers into a voting system could achieve better performance than a single classifier in recognizing medication information from clinical text. It suggests that simple strategies that can be easily implemented such as majority voting could have the potential to significantly improve clinical entity recognition.
We describe Support Vector Machine (SVM) applications to classification and clustering of channel current data. SVMs are variational-calculus based methods that are constrained to have structural risk minimization (SRM), i.e., they provide noise tolerant solutions for pattern recognition. The SVM approach encapsulates a significant amount of model-fitting information in the choice of its kernel. In work thus far, novel, information-theoretic, kernels have been successfully employed for notably better performance over standard kernels. Currently there are two approaches for implementing multiclass SVMs. One is called external multi-class that arranges several binary classifiers as a decision tree such that they perform a single-class decision making function, with each leaf corresponding to a unique class. The second approach, namely internal-multiclass, involves solving a single optimization problem corresponding to the entire data set (with multiple hyperplanes).
Each SVM approach encapsulates a significant amount of model-fitting information in its choice of kernel. In work thus far, novel, information-theoretic, kernels were successfully employed for notably better performance over standard kernels. Two SVM approaches to multiclass discrimination are described: (1) internal multiclass (with a single optimization), and (2) external multiclass (using an optimized decision tree). We describe benefits of the internal-SVM approach, along with further refinements to the internal-multiclass SVM algorithms that offer significant improvement in training time without sacrificing accuracy. In situations where the data isn't clearly separable, making for poor discrimination, signal clustering is used to provide robust and useful information – to this end, novel, SVM-based clustering methods are also described. As with the classification, there are Internal and External SVM Clustering algorithms, both of which are briefly described.
Biomedical texts can be typically represented by four rhetorical categories: Introduction, Methods, Results and Discussion (IMRAD). Classifying sentences into these categories can benefit many other text-mining tasks. Although many studies have applied different approaches for automatically classifying sentences in MEDLINE abstracts into the IMRAD categories, few have explored the classification of sentences that appear in full-text biomedical articles. We first evaluated whether sentences in full-text biomedical articles could be reliably annotated into the IMRAD format and then explored different approaches for automatically classifying these sentences into the IMRAD categories. Our results show an overall annotation agreement of 82.14% with a Kappa score of 0.756. The best classification system is a multinomial naïve Bayes classifier trained on manually annotated data that achieved 91.95% accuracy and an average F-score of 91.55%, which is significantly higher than baseline systems. A web version of this system is available online at—http://wood.ims.uwm.edu/full_text_classifier/.
Protein remote homology detection and fold recognition are central problems in computational biology. Supervised learning algorithms based on support vector machines are currently one of the most effective methods for solving these problems. These methods are primarily used to solve binary classification problems and they have not been extensively used to solve the more general multiclass remote homology prediction and fold recognition problems.
We present a comprehensive evaluation of a number of methods for building SVM-based multiclass classification schemes in the context of the SCOP protein classification. These methods include schemes that directly build an SVM-based multiclass model, schemes that employ a second-level learning approach to combine the predictions generated by a set of binary SVM-based classifiers, and schemes that build and combine binary classifiers for various levels of the SCOP hierarchy beyond those defining the target classes.
Analyzing the performance achieved by the different approaches on four different datasets we show that most of the proposed multiclass SVM-based classification approaches are quite effective in solving the remote homology prediction and fold recognition problems and that the schemes that use predictions from binary models constructed for ancestral categories within the SCOP hierarchy tend to not only lead to lower error rates but also reduce the number of errors in which a superfamily is assigned to an entirely different fold and a fold is predicted as being from a different SCOP class. Our results also show that the limited size of the training data makes it hard to learn complex second-level models, and that models of moderate complexity lead to consistently better results.
This paper describes a system developed for Track 2 of the 2011 Medical NLP Challenge on identifying emotions in suicide notes. Our approach involves learning a collection of one-versus-all classifiers, each deciding whether or not a particular label should be assigned to a given sentence. We explore a variety of features types—syntactic, semantic and surface-oriented. Cost-sensitive learning is used for dealing with the issue of class imbalance in the data.
emotion classification; suicidology; support vector machines; cost-sensitive learning
Protein subcellular localization is crucial for genome annotation, protein function prediction, and drug discovery. Determination of subcellular localization using experimental approaches is time-consuming; thus, computational approaches become highly desirable. Extensive studies of localization prediction have led to the development of several methods including composition-based and homology-based methods. However, their performance might be significantly degraded if homologous sequences are not detected. Moreover, methods that integrate various features could suffer from the problem of low coverage in high-throughput proteomic analyses due to the lack of information to characterize unknown proteins.
We propose a hybrid prediction method for Gram-negative bacteria that combines a one-versus-one support vector machines (SVM) model and a structural homology approach. The SVM model comprises a number of binary classifiers, in which biological features derived from Gram-negative bacteria translocation pathways are incorporated. In the structural homology approach, we employ secondary structure alignment for structural similarity comparison and assign the known localization of the top-ranked protein as the predicted localization of a query protein. The hybrid method achieves overall accuracy of 93.7% and 93.2% using ten-fold cross-validation on the benchmark data sets. In the assessment of the evaluation data sets, our method also attains accurate prediction accuracy of 84.0%, especially when testing on sequences with a low level of homology to the training data. A three-way data split procedure is also incorporated to prevent overestimation of the predictive performance. In addition, we show that the prediction accuracy should be approximately 85% for non-redundant data sets of sequence identity less than 30%.
Our results demonstrate that biological features derived from Gram-negative bacteria translocation pathways yield a significant improvement. The biological features are interpretable and can be applied in advanced analyses and experimental designs. Moreover, the overall accuracy of combining the structural homology approach is further improved, which suggests that structural conservation could be a useful indicator for inferring localization in addition to sequence homology. The proposed method can be used in large-scale analyses of proteomes.
We investigate the problems of multiclass cancer classification
with gene selection from gene expression data. Two different
constructed multiclass classifiers with gene selection are
proposed, which are fuzzy support vector machine (FSVM) with gene
selection and binary classification tree based on SVM with gene
selection. Using F test and recursive feature elimination based on
SVM as gene selection methods, binary classification tree based on
SVM with F test, binary classification tree based on SVM with
recursive feature elimination based on SVM, and FSVM with
recursive feature elimination based on SVM are tested in our
experiments. To accelerate computation, preselecting the strongest
genes is also used. The proposed techniques are applied to analyze
breast cancer data, small round blue-cell tumors, and acute
leukemia data. Compared to existing multiclass cancer classifiers
and binary classification tree based on SVM with F test or binary
classification tree based on SVM with recursive feature
elimination based on SVM mentioned in this paper, FSVM based on
recursive feature elimination based on SVM can find most important
genes that affect certain types of cancer with high recognition accuracy.
Pattern recognition approaches, such as the Support Vector Machine (SVM), have been successfully used to classify groups of individuals based on their patterns of brain activity or structure. However these approaches focus on finding group differences and are not applicable to situations where one is interested in accessing deviations from a specific class or population. In the present work we propose an application of the one-class SVM (OC-SVM) to investigate if patterns of fMRI response to sad facial expressions in depressed patients would be classified as outliers in relation to patterns of healthy control subjects. We defined features based on whole brain voxels and anatomical regions. In both cases we found a significant correlation between the OC-SVM predictions and the patients' Hamilton Rating Scale for Depression (HRSD), i.e. the more depressed the patients were the more of an outlier they were. In addition the OC-SVM split the patient groups into two subgroups whose membership was associated with future response to treatment. When applied to region-based features the OC-SVM classified 52% of patients as outliers. However among the patients classified as outliers 70% did not respond to treatment and among those classified as non-outliers 89% responded to treatment. In addition 89% of the healthy controls were classified as non-outliers.
► It is possible to define a “normality boundary” based on patterns of fMRI from healthy subjects. ► Most severely depressed patients should lie outside this boundary. ► The amount of departure from the “normality boundary” is related to the severity of the depression.
fMRI; Pattern classification; Depression; Machine learning; Support Vector Machine; Outlier detection
Building classifiers for medical problems often involves dealing with rare, but important events. Imbalanced datasets pose challenges to ordinary classification algorithms such as Logistic Regression (LR) and Support Vector Machines (SVM). The lack of effective strategies for dealing with imbalanced training data often results in models that exhibit poor discrimination. We propose a novel approach to estimate class memberships based on the evaluation of pairwise relationships in the training data. The method we propose, Pairwise Expanded Logistic Regression, improved discrimination and had higher accuracy when compared to existing methods in two imbalanced datasets, thus showing promise as a potential remedy for this problem.
In this paper we report on the approaches that we developed for the 2011 i2b2 Shared Task on Sentiment Analysis of Suicide Notes. We have cast the problem of detecting emotions in suicide notes as a supervised multi-label classification problem. Our classifiers use a variety of features based on (a) lexical indicators, (b) topic scores, and (c) similarity measures. Our best submission has a precision of 0.551, a recall of 0.485, and a F-measure of 0.516.
similarity method; statistical method; sentiment classification; suicide notes
Using a classification scheme of patient medication status we sought to recognize and categorize medications mentioned in the unrestricted text of clinical documents generated in clinical practice. The categories refer to the patient's status with respect to the medication such as discontinuation, start or initiation, and continuation of a given medication. This categorization is performed with a machine learning technique, Maximum Entropy (ME), that is well suited to incorporating heterogeneous sources of information necessary for classifying patient's medication status. We use hand labeled training data to generate ME models and test 5 different training feature sets. Our results show that the most optimal feature set includes a combination of the following: two words preceding and following the mention of the drug, the subject of the sentence in which the drug mention occurs, the 2 words following the subject, and a binary feature vector of lexicalized semantic cues indicative of medication status or its change. The average predictive power of a model trained on these features is approximately 89%.