List of usage examples for edu.stanford.nlp.trees TypedDependency dep
IndexedWord dep
To view the source code for edu.stanford.nlp.trees TypedDependency dep.
Click Source Link
From source file:DependencyParse.java
License:Apache License
public static void main(String[] args) throws Exception { Properties props = StringUtils.argsToProperties(args); if (!props.containsKey("tokpath") || !props.containsKey("parentpath") || !props.containsKey("relpath")) { System.err.println(// w ww . jav a2s .com "usage: java DependencyParse -tokenize - -tokpath <tokpath> -parentpath <parentpath> -relpath <relpath>"); System.exit(1); } boolean tokenize = false; if (props.containsKey("tokenize")) { tokenize = true; } String tokPath = props.getProperty("tokpath"); String parentPath = props.getProperty("parentpath"); String relPath = props.getProperty("relpath"); BufferedWriter tokWriter = new BufferedWriter(new FileWriter(tokPath)); BufferedWriter parentWriter = new BufferedWriter(new FileWriter(parentPath)); BufferedWriter relWriter = new BufferedWriter(new FileWriter(relPath)); MaxentTagger tagger = new MaxentTagger(TAGGER_MODEL); DependencyParser parser = DependencyParser.loadFromModelFile(PARSER_MODEL); Scanner stdin = new Scanner(System.in); int count = 0; long start = System.currentTimeMillis(); while (stdin.hasNextLine()) { String line = stdin.nextLine(); List<HasWord> tokens = new ArrayList<>(); if (tokenize) { PTBTokenizer<Word> tokenizer = new PTBTokenizer(new StringReader(line), new WordTokenFactory(), ""); for (Word label; tokenizer.hasNext();) { tokens.add(tokenizer.next()); } } else { for (String word : line.split(" ")) { tokens.add(new Word(word)); } } List<TaggedWord> tagged = tagger.tagSentence(tokens); int len = tagged.size(); Collection<TypedDependency> tdl = parser.predict(tagged).typedDependencies(); int[] parents = new int[len]; for (int i = 0; i < len; i++) { // if a node has a parent of -1 at the end of parsing, then the node // has no parent. parents[i] = -1; } String[] relns = new String[len]; for (TypedDependency td : tdl) { // let root have index 0 int child = td.dep().index(); int parent = td.gov().index(); relns[child - 1] = td.reln().toString(); parents[child - 1] = parent; } // print tokens StringBuilder sb = new StringBuilder(); for (int i = 0; i < len - 1; i++) { if (tokenize) { sb.append(PTBTokenizer.ptbToken2Text(tokens.get(i).word())); } else { sb.append(tokens.get(i).word()); } sb.append(' '); } if (tokenize) { sb.append(PTBTokenizer.ptbToken2Text(tokens.get(len - 1).word())); } else { sb.append(tokens.get(len - 1).word()); } sb.append('\n'); tokWriter.write(sb.toString()); // print parent pointers sb = new StringBuilder(); for (int i = 0; i < len - 1; i++) { sb.append(parents[i]); sb.append(' '); } sb.append(parents[len - 1]); sb.append('\n'); parentWriter.write(sb.toString()); // print relations sb = new StringBuilder(); for (int i = 0; i < len - 1; i++) { sb.append(relns[i]); sb.append(' '); } sb.append(relns[len - 1]); sb.append('\n'); relWriter.write(sb.toString()); count++; if (count % 1000 == 0) { double elapsed = (System.currentTimeMillis() - start) / 1000.0; System.err.printf("Parsed %d lines (%.2fs)\n", count, elapsed); } } long totalTimeMillis = System.currentTimeMillis() - start; System.err.printf("Done: %d lines in %.2fs (%.1fms per line)\n", count, totalTimeMillis / 1000.0, totalTimeMillis / (double) count); tokWriter.close(); parentWriter.close(); relWriter.close(); }
From source file:ConstituencyParse.java
License:Apache License
public int[] depTreeParents(Tree tree, List<HasWord> tokens) { GrammaticalStructure gs = gsf.newGrammaticalStructure(tree); Collection<TypedDependency> tdl = gs.typedDependencies(); int len = tokens.size(); int[] parents = new int[len]; for (int i = 0; i < len; i++) { // if a node has a parent of -1 at the end of parsing, then the node // has no parent. parents[i] = -1;// w w w . ja va2 s .co m } for (TypedDependency td : tdl) { // let root have index 0 int child = td.dep().index(); int parent = td.gov().index(); parents[child - 1] = parent; } return parents; }
From source file:ca.ualberta.exemplar.core.ParserMalt.java
License:Open Source License
@Override public List<CoreMap> parseText(String text) { List<CoreMap> sentences = null; try {//w w w. java 2s . com Annotation document = new Annotation(text); pipeline.annotate(document); sentences = document.get(SentencesAnnotation.class); for (CoreMap sentence : sentences) { List<CoreLabel> tokens = sentence.get(TokensAnnotation.class); String[] conllInput = sentenceToCoNLLInput(tokens); DependencyGraph graph = (DependencyGraph) maltParser.parse(conllInput); Result result = graphToCoNLL(graph); List<List<String>> conll = result.conll; int rootIndex = result.rootIndex; EnglishGrammaticalStructure gs = EnglishGrammaticalStructure.buildCoNNLXGrammaticStructure(conll); TreeGraphNode root = null; List<TypedDependency> deps = gs.typedDependenciesCCprocessed(); // Add root dependency and ner annotations int size = deps.size(); for (int i = 0; i < size; i++) { TypedDependency td = deps.get(i); if (td.gov().index() == rootIndex) { root = td.gov(); deps.add(new TypedDependency(GrammaticalRelation.ROOT, td.gov(), td.gov())); } { TreeGraphNode n = td.dep(); if (n.label().ner() == null) { n.label().setNER(tokens.get(n.index() - 1).ner()); n.label().setBeginPosition(tokens.get(n.index() - 1).beginPosition()); n.label().setEndPosition(tokens.get(n.index() - 1).endPosition()); n.label().setLemma(tokens.get(n.index() - 1).lemma()); } } { TreeGraphNode n = td.gov(); if (n.label().ner() == null) { n.label().setNER(tokens.get(n.index() - 1).ner()); n.label().setBeginPosition(tokens.get(n.index() - 1).beginPosition()); n.label().setEndPosition(tokens.get(n.index() - 1).endPosition()); n.label().setLemma(tokens.get(n.index() - 1).lemma()); } } } if (root == null) continue; List<TreeGraphNode> roots = new ArrayList<TreeGraphNode>(); roots.add(gs.root()); SemanticGraph sg = new SemanticGraph(deps, roots); sentence.set(CollapsedCCProcessedDependenciesAnnotation.class, sg); } } catch (Exception e) { e.printStackTrace(); } return sentences; }
From source file:cc.vidr.parseviz.ParseViz.java
License:Open Source License
public static void printDependenciesDot(Tree tree, StringBuilder sb) { sb.append("digraph{\n"); for (TypedDependency td : typedDependencies(tree)) { GrammaticalRelation reln = td.reln(); TreeGraphNode gov = td.gov();/*from w w w .ja va2 s . c o m*/ TreeGraphNode dep = td.dep(); sb.append("n").append(gov.index()).append("[label=\"").append(gov).append("\"];\n").append("n") .append(dep.index()).append("[label=\"").append(dep).append("\"];\n").append("n") .append(gov.index()).append("->n").append(dep.index()).append("[label=\"").append(reln) .append("\"];\n"); } sb.append("}\n"); }
From source file:com.epictodo.controller.nlp.NLPEngine.java
License:Open Source License
/** * This method understands the natural input from an input sentence * The sentence is analyzed and important information will be extracted and returned * <p/>//w w w . j a va 2 s. c o m * Assumptions: * 1. PERSON name has to start with a capital letter. For example, "Kenneth" * 2. Priority is determined by weekly basis from today's date. Priority increases every week. * <p/> * 3. Time has to strictly follow the format (hh:mm) * 3.1 10:00 is accepted * 3.2 1000 is not accepted * <p/> * 4. PERSON name cannot be named after a month (for example, April, May, June) -> Registers as a DATE * 5. PERSON name cannot be named after a day (for example, Sunday) -> Registers as a DATE * <p/> * 6. Time period: Day cannot be placed after the time period * 6.1. Tuesday 2 weeks later (PASSED) - gets the Tuesday 2 weeks later * 6.2. 2 weeks later Tuesday (FAILED) - gets 2 weeks later from today's date * <p/> * 7. There can be only 2 time input which will be allowed to be parsed in. Additional time will be rejected. * 7.1. from 10:00 to 14:00 then 18:00 > reflects only [10:00, 14:00] as start_time and end_time * <p/> * 8. If users input date in quick date for example, (01/12/2014 -> dd/MM/yyyy) * 8.1. [Issue] Manual correction of storing through validation * <p/> * 9. Chinese names will be read as two different PERSON. For example, Jie Ning will be read as 'Jie', 'Ning' * 10. Supports one short date, for example '14/11/2014' * <p/> * Usage: * <p/> * 1. flexiAdd("meeting with Damith on project submission Tuesday 2 weeks later at 10:34"); * 2. flexiAdd("project submission for cs2103 next Tuesday from 10:00 to 14:00 to Damith"); * <p/> * Output: * <p/> * 1. * Task Name: meeting submission * Task Desc: Damith Tuesday 2 weeks later at 10:34 * Task Date: 181114 * Task Time: 10:34 * Task Priority: 8 * Task Start Time: null * Task End Time: null * Task Duration: 0.0 * <p/> * 2. * Task Name: submission cs2103 * Task Desc: Damith 10:00 14:00 next Tuesday * Task Date: 111114 * Task Time: null * Task Priority: 9 * Task Start Time: 10:00 * Task End Time: 14:00 * Task Duration: 4.0 * * @param _sentence * @throws ParseException */ public Response flexiAdd(String _sentence) throws ParseException { _response = new Response(); boolean is_date_set = false; boolean is_time_set = false; boolean is_priority_set = false; String tomorrow_date; String date_value; String time_value; String _priority; String start_time; String end_time; double _duration; int num_days; List<String> analyzed_results = new ArrayList<>(); List<String> task_name = new ArrayList<>(); List<String> task_desc = new ArrayList<>(); List<String> task_date = new ArrayList<>(); List<String> task_time = new ArrayList<>(); List<String> date_list = new ArrayList<>(); StringBuilder string_builder = new StringBuilder(CAPACITY); // Initialize Sentence Structure & Sentence Analysis to Map Map<String, String> sentence_struct_map = sentence_struct.sentenceDependencies(_sentence); Map<String, String> date_time_map = sentence_analysis.dateTimeAnalyzer(_sentence); Map<String, String> sentence_token_map = sentence_analysis.sentenceAnalyzer(_sentence); LinkedHashMap<String, LinkedHashSet<String>> entities_map = sentence_analysis .nerEntitiesExtractor(_sentence); List<TypedDependency> grammar_struct_map = grammartical_parser.grammarAnalyzer(_sentence); /** * Name Entity Recognition (NER) map * For loop to traverse key:value of NER map to retrieve keywords to be processed * * Example: root, nn, dobj, nsubj, aux, xcomp, prep, num, etc. * */ for (Map.Entry<String, LinkedHashSet<String>> map_result : entities_map.entrySet()) { String _key = map_result.getKey(); LinkedHashSet<String> _value = map_result.getValue(); } /** * Sentence analyzer to analyze the text for keywords * For loop to traverse the following: * * 1. (TIME) - stores TIME's (value) key to check if it's a single or multiple TIME value * 1.1. Checks if key is valid time before storing. * 1.2. "at" maybe classified as a valid TIME but we do not want to register this * * 2. (PERSON) - stores PERSON name to be processed in Task Description * 2.1. PERSON does not accept 'Prof' as an actual person, rather an entity * * 3. (LOCATION) - stores LOCATION name if it exists. * 3.1. TOKYO, SINGAPORE are actual LOCATION * 3.2. COM1, LT15, SR1 are not an actual LOCATION stored in our model (can be implemented using Classifier in future) * */ for (Map.Entry<String, String> map_result : sentence_token_map.entrySet()) { String _key = map_result.getKey(); String _value = map_result.getValue(); if (_value.equalsIgnoreCase("TIME")) { if (time_validator.validate(_key)) { task_time.add(_key); } } if (_value.equalsIgnoreCase("NUMBER")) { if (date_validator.validateDateExpression(_key)) { date_value = date_validator.fixShortDate(_key); task_date.add(date_value); num_days = date_validator.compareDate(date_value); // Checks if date distance is >= 0 or <= 1 of 1 day if (num_days >= 0 && num_days <= 1) { _priority = date_validator.determinePriority(date_value); date_value = date_validator.genericDateFormat(date_value); if (_response.getTaskDate() == null) { _response.setTaskDate(date_value); } _response.setPriority(Integer.parseInt(_priority)); is_priority_set = true; } else { // Check if TaskDate has been set previously, prevent override _priority = date_validator.determinePriority(date_value); date_value = date_validator.genericDateFormat(date_value); if (_response.getTaskDate() == null) { _response.setTaskDate(date_value); } _response.setPriority(Integer.parseInt(_priority)); is_priority_set = true; } is_date_set = true; } else { date_value = date_validator.nlpShortDate(_key); task_date.add(date_value); num_days = date_validator.compareDate(date_value); // Checks if date distance is >= 0 or <= 1 of 1 day if (num_days >= 0 && num_days <= 1) { _priority = date_validator.determinePriority(date_value); date_value = date_validator.genericDateFormat(date_value); if (_response.getTaskDate() == null) { _response.setTaskDate(date_value); } _response.setPriority(Integer.parseInt(_priority)); is_priority_set = true; } else { // Check if TaskDate has been set previously, prevent override _priority = date_validator.determinePriority(date_value); date_value = date_validator.genericDateFormat(date_value); if (_response.getTaskDate() == null) { _response.setTaskDate(date_value); } _response.setPriority(Integer.parseInt(_priority)); is_priority_set = true; } is_date_set = true; } } if (_value.equalsIgnoreCase("PERSON")) { if (!_key.equalsIgnoreCase("Prof")) { task_desc.add(_key); analyzed_results.add(_key); } else { task_desc.add(_key); analyzed_results.add(_key); } } else if (_value.equalsIgnoreCase("LOCATION")) { task_desc.add(_key); analyzed_results.add(_key); } else if (_value.equalsIgnoreCase("ORGANIZATION")) { task_desc.add(_key); analyzed_results.add(_key); } } /** * This algorithm checks if the time values stored are more than or equals to 2 * There can be instances where users input 3 or more time variables, but the first 2 will be registered * * This algorithm will getTimeDuration before storing start_time, end_time & _duration to _response */ if (task_time.size() != 0 && task_time.size() >= 2) { start_time = task_time.get(0); end_time = task_time.get(1); _duration = time_validator.getTimeDuration(start_time, end_time); _response.setStartTime(start_time); _response.setEndTime(end_time); _response.setTaskDuration(_duration); is_time_set = true; } else if (task_time.size() != 0 && task_time.size() < 2) { _response.setTaskTime(task_time.get(0)); is_time_set = true; } /** * Date time analyzer map to analyze date time values to be stored * This algorithm will check for TOMORROW date & time distance. * * If that case happens, the result will be parsed to getDateInFormat & getTimeInFormat to handle such cases * Otherwise, the result will be parsed to validateDate & validateTime to handle the more generic cases */ for (Map.Entry<String, String> map_result : date_time_map.entrySet()) { String _key = map_result.getKey(); String _value = map_result.getValue(); if (!is_date_set) { tomorrow_date = date_validator.convertDateFormat(_value); num_days = date_validator.compareDate(tomorrow_date); // Checks if date distance is >= 0 or <= 1 of 1 day if (num_days >= 0 && num_days <= 1) { _priority = date_validator.determinePriority(tomorrow_date); date_value = date_validator.genericDateFormat(tomorrow_date); time_value = date_validator.getTimeInFormat(_value); if (!is_date_set) { // Check if TaskDate has been set previously, prevent override if (date_value.length() == 5) { date_list.add(0, "0"); Scanner _scanner = new Scanner(date_value); while (_scanner.hasNext()) { date_list.add(_scanner.next()); } _scanner.close(); for (String date : date_list) { string_builder.append(date); } _response.setTaskDate(string_builder.toString()); is_date_set = true; } else { _response.setTaskDate(date_value); is_date_set = true; } } if (!is_time_set) { _response.setTaskTime(time_value); is_time_set = true; } if (!is_priority_set) { _response.setPriority(Integer.parseInt(_priority)); is_priority_set = true; } // _response.setTaskDate(date_value); // _response.setTaskTime(time_value); // _response.setPriority(Integer.parseInt(_priority)); task_desc.add(_key); analyzed_results.add(_key); analyzed_results.add(date_value); analyzed_results.add(time_value); analyzed_results.add(_priority); } else { // Check if TaskDate has been set previously, prevent override _priority = date_validator.determinePriority(_value); date_value = date_validator.genericDateFormat(tomorrow_date); time_value = date_validator.validateTime(_value); if (!is_date_set) { if (date_value.length() == 5) { date_list.add(0, "0"); Scanner _scanner = new Scanner(date_value); while (_scanner.hasNext()) { date_list.add(_scanner.next()); } _scanner.close(); for (String date : date_list) { string_builder.append(date); } _response.setTaskDate(string_builder.toString()); is_date_set = true; } else { _response.setTaskDate(date_value); is_date_set = true; } } if (!is_time_set) { _response.setTaskTime(time_value); is_time_set = true; } if (!is_priority_set) { _response.setPriority(Integer.parseInt(_priority)); is_priority_set = true; } // _response.setTaskDate(date_value); // _response.setTaskTime(time_value); // _response.setPriority(Integer.parseInt(_priority)); task_desc.add(_key); analyzed_results.add(_key); analyzed_results.add(date_value); analyzed_results.add(time_value); analyzed_results.add(_priority); } } } /** * Sentence Dependencies map to analyze and return the dependencies of the sentence structure * This algorithm checks the dependencies relationship in the tree structure, and returns the results */ for (Map.Entry<String, String> map_result : sentence_struct_map.entrySet()) { String _key = map_result.getKey(); String _value = map_result.getValue(); if ((_key.equalsIgnoreCase("root") || _key.equalsIgnoreCase("dep") || _key.equalsIgnoreCase("dobj") || _key.equalsIgnoreCase("prep_on") || _key.equalsIgnoreCase("prep_for") || _key.equalsIgnoreCase("nn") || _key.equalsIgnoreCase("xcomp")) && (_value.equalsIgnoreCase("aux"))) { task_name.add(_value); analyzed_results.add(_value); } } /** * Grammartical Struct map analyzes and return the relationship and dependencies of the grammartical structure * This algorithm checks for * * 1. root: the grammatical relation that points to the root of the sentence * 1.1. Remove words like 'be' * 1.2. Remove words that are already added into the list * 2. nn: noun compound modifier is any noun that serves to modify the head noun * 2.1. Remove words that is stored already in Task Description * 3. nsubj: nominal subject is a noun phrase which is the syntactic subject of a clause * aux: auxiliary is a non-main verb of the clause * xcomp: open clausal complement is a clausal complement without * its own subject, whose reference is determined by an external subject * dobj: direct object is the noun phrase which is the (accusative) object of the verb * 3.1. Add words that does not already exist in Task Name * 4. amod: adjectival modifier is any adjectival phrase that serves to modify the meaning of the NP * 4.1. Remove words like 'next' * 5. prep: prepositional modifier of a verb, adjective, or noun is any prepositional phrase that serves to * modify the meaning of the verb, adjective, noun, or even another prepositon * 5.1. Check Task Time size is more than 1 or less than 2 * 5.2. Remove Task Time accordingly * 6. dep: dependent is when the system is unable to determine a more precise * dependency relation between two words * 6.1. Remove words like 'regarding' * 7. aux: is a non-main verb of the clause, e.g., a modal auxiliary, or a form of * "be", "do" or "have" in a periphrastic tense * 7.1. Remove words like 'to' * * This algorithm ensures that NLP maintains a certain integrity to making sense while parsing the sentence * */ for (TypedDependency type_dependency : grammar_struct_map) { String reln_key = type_dependency.reln().getShortName(); String dep_value = type_dependency.dep().nodeString(); if (reln_key.equalsIgnoreCase("root")) { if (task_name.contains("Be") || task_name.contains("be")) { task_name.remove(dep_value); } else if (task_name.contains(dep_value)) { task_name.remove(dep_value); } else { task_name.add(dep_value); } } if (reln_key.equalsIgnoreCase("nn")) { if (!task_name.contains(dep_value)) { task_name.add(dep_value); } if (task_desc.contains(dep_value)) { task_name.remove(dep_value); } } if (reln_key.equalsIgnoreCase("nsubj") || reln_key.equalsIgnoreCase("aux") || reln_key.equalsIgnoreCase("xcomp") || reln_key.equalsIgnoreCase("dobj")) { if (!task_name.contains(dep_value)) { task_name.add(dep_value); } } if (reln_key.equalsIgnoreCase("amod") && !dep_value.equalsIgnoreCase("next")) { if (!task_name.contains(dep_value)) { task_name.add(dep_value); } } if (reln_key.equalsIgnoreCase("prep")) { if (!task_name.contains(dep_value) && (!date_validator.checkDateFormat(dep_value) && !date_validator.checkDateFormat2(dep_value))) { task_name.add(dep_value); } if (task_time.size() > 0 && task_time.size() < 2) { if (dep_value.contains(task_time.get(0))) { task_name.remove(dep_value); } } else if (task_time.size() >= 2 && task_time.size() < 3) { if (dep_value.contains(task_time.get(0)) || dep_value.contains(task_time.get(1))) { task_name.remove(dep_value); } } } if (reln_key.equalsIgnoreCase("dep") && dep_value.equalsIgnoreCase("regarding")) { if (task_name.contains(dep_value)) { task_name.remove(dep_value); } } if (reln_key.equalsIgnoreCase("aux") && dep_value.equalsIgnoreCase("to")) { if (task_name.contains(dep_value)) { task_name.remove(dep_value); } } } _response.setTaskName(task_name); _response.setTaskDesc(task_desc); return _response; }
From source file:com.search.MySearchHandler.java
License:Apache License
private static String compress(TermTokenStream queryStream) { List<TypedDependency> tdl = getTyped(queryStream.toNLPString()); ListIterator<TypedDependency> tdllist = tdl.listIterator(); int count = 1; while (tdllist.hasNext()) { TypedDependency typd = tdllist.next(); // Have to fix the multiple words in the sequence issue if ((typd.reln().toString().equals("nn") | typd.reln().toString().equals("amod") | (typd.reln().toString().equals("det") && typd.dep().index() != 1) | typd.reln().toString().equals("num") | typd.reln().toString().equals("number")) && ((typd.gov().index() - typd.dep().index()) == 1 || (typd.gov().index() - typd.dep().index()) == -1)) { if ((typd.gov().index() - typd.dep().index()) == -1) { queryStream.mergeWithNext(typd.gov().index() - count); count++;/* w ww . j ava 2 s. co m*/ } else { queryStream.mergeWithNext(typd.dep().index() - count); count++; } } } if (found(getTyped(queryStream.toNLPString()))) { return compress(queryStream); } return queryStream.toNLPString(); }
From source file:com.search.MySearchHandler.java
License:Apache License
private static boolean found(List<TypedDependency> typed) { ListIterator<TypedDependency> tdllist = typed.listIterator(); boolean iftrue = false; while (tdllist.hasNext()) { TypedDependency typd = tdllist.next(); // Have to fix the multiple words in the sequence issue if ((typd.reln().toString().equals("nn") | typd.reln().toString().equals("amod") | (typd.reln().toString().equals("det") && typd.dep().index() != 1) | typd.reln().toString().equals("num") | typd.reln().toString().equals("number")) && ((typd.gov().index() - typd.dep().index()) == 1 || (typd.gov().index() - typd.dep().index()) == -1)) { iftrue = true;// w w w . ja v a 2 s . c o m } } return iftrue; }
From source file:coreferenceresolver.element.Token.java
/** * @param newTokenSentiment the sentimentOrientation to set *//*from w w w. j a va 2 s . co m*/ public void setSentimentOrientation(int newTokenSentiment, Collection<TypedDependency> typedDeps) { if ((newTokenSentiment == Util.POSITIVE || newTokenSentiment == Util.NEGATIVE) && (!this.POS.equals("IN"))) { for (TypedDependency typedDependency : typedDeps) { if (typedDependency.reln().toString().equals("neg") && typedDependency.gov().value().equals(word)) { newTokenSentiment = Util.reverseSentiment(newTokenSentiment); } if (typedDependency.reln().toString().equals("mark") && (typedDependency.dep().value().toLowerCase().equals("although") || typedDependency.dep().value().toLowerCase().equals("though") || typedDependency.dep().value().toLowerCase().equals("while")) && typedDependency.gov().value().equals(word)) { newTokenSentiment = Util.NEUTRAL; break; } } } this.sentimentOrientation = newTokenSentiment; }
From source file:de.tudarmstadt.ukp.dkpro.core.stanfordnlp.StanfordDependencyConverter.java
License:Open Source License
protected void doCreateDependencyTags(JCas aJCas, TreebankLanguagePack aLP, Tree parseTree, List<Token> tokens) { GrammaticalStructure gs;/*from w w w . j a v a2 s. c o m*/ try { gs = aLP.grammaticalStructureFactory(aLP.punctuationWordRejectFilter(), aLP.typedDependencyHeadFinder()) .newGrammaticalStructure(parseTree); } catch (UnsupportedOperationException e) { // We already warned in the model provider if dependencies are not supported, so here // we just do nothing and skip the dependencies. return; } Collection<TypedDependency> dependencies = null; switch (mode) { case BASIC: dependencies = gs.typedDependencies(); // gs.typedDependencies(false); break; case NON_COLLAPSED: dependencies = gs.allTypedDependencies(); // gs.typedDependencies(true); break; case COLLAPSED_WITH_EXTRA: dependencies = gs.typedDependenciesCollapsed(true); break; case COLLAPSED: dependencies = gs.typedDependenciesCollapsed(false); break; case CC_PROPAGATED: dependencies = gs.typedDependenciesCCprocessed(true); break; case CC_PROPAGATED_NO_EXTRA: dependencies = gs.typedDependenciesCCprocessed(false); break; case TREE: dependencies = gs.typedDependenciesCollapsedTree(); break; } for (TypedDependency currTypedDep : dependencies) { int govIndex = currTypedDep.gov().index(); int depIndex = currTypedDep.dep().index(); if (govIndex != 0) { // Stanford CoreNLP produces a dependency relation between a verb and ROOT-0 which // is not token at all! Token govToken = tokens.get(govIndex - 1); Token depToken = tokens.get(depIndex - 1); StanfordAnnotator.createDependencyAnnotation(aJCas, currTypedDep.reln(), govToken, depToken); } } }
From source file:de.tudarmstadt.ukp.dkpro.core.stanfordnlp.StanfordParser.java
License:Open Source License
protected void doCreateDependencyTags(ParserGrammar aParser, StanfordAnnotator sfAnnotator, Tree parseTree, List<Token> tokens) { GrammaticalStructure gs;//from w w w.j a v a 2s. c o m try { TreebankLanguagePack tlp = aParser.getTLPParams().treebankLanguagePack(); gs = tlp.grammaticalStructureFactory(tlp.punctuationWordRejectFilter(), tlp.typedDependencyHeadFinder()) .newGrammaticalStructure(parseTree); } catch (UnsupportedOperationException e) { // We already warned in the model provider if dependencies are not supported, so here // we just do nothing and skip the dependencies. return; } Collection<TypedDependency> dependencies = null; switch (mode) { case BASIC: dependencies = gs.typedDependencies(); // gs.typedDependencies(false); break; case NON_COLLAPSED: dependencies = gs.allTypedDependencies(); // gs.typedDependencies(true); break; case COLLAPSED_WITH_EXTRA: dependencies = gs.typedDependenciesCollapsed(true); break; case COLLAPSED: dependencies = gs.typedDependenciesCollapsed(false); break; case CC_PROPAGATED: dependencies = gs.typedDependenciesCCprocessed(true); break; case CC_PROPAGATED_NO_EXTRA: dependencies = gs.typedDependenciesCCprocessed(false); break; case TREE: dependencies = gs.typedDependenciesCollapsedTree(); break; } for (TypedDependency currTypedDep : dependencies) { int govIndex = currTypedDep.gov().index(); int depIndex = currTypedDep.dep().index(); if (govIndex != 0) { // Stanford CoreNLP produces a dependency relation between a verb and ROOT-0 which // is not token at all! Token govToken = tokens.get(govIndex - 1); Token depToken = tokens.get(depIndex - 1); sfAnnotator.createDependencyAnnotation(currTypedDep.reln(), govToken, depToken); } } }