-
Notifications
You must be signed in to change notification settings - Fork 288
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Add wordnet based lemmatizer (#1991)
### What problem does this PR solve? Remove libwn, because it could not return same results as NLTK.WordNetLemmatizer Issue link:#1973 ### Type of change - [X] New Feature (non-breaking change which adds functionality)
- Loading branch information
Showing
16 changed files
with
292 additions
and
5,159 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,219 @@ | ||
| // Copyright(C) 2024 InfiniFlow, Inc. All rights reserved. | ||
| // | ||
| // Licensed under the Apache License, Version 2.0 (the "License"); | ||
| // you may not use this file except in compliance with the License. | ||
| // You may obtain a copy of the License at | ||
| // | ||
| // https://www.apache.org/licenses/LICENSE-2.0 | ||
| // | ||
| // Unless required by applicable law or agreed to in writing, software | ||
| // distributed under the License is distributed on an "AS IS" BASIS, | ||
| // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| // See the License for the specific language governing permissions and | ||
| // limitations under the License. | ||
| module; | ||
|
|
||
| #include <cassert> | ||
| #include <cstddef> | ||
| #include <fstream> | ||
| #include <iostream> | ||
| #include <sstream> | ||
| #include <unordered_map> | ||
|
|
||
| module lemmatizer; | ||
|
|
||
| import stl; | ||
| import third_party; | ||
|
|
||
| namespace infinity { | ||
|
|
||
| static const String ADJ = "a"; | ||
| static const String ADJ_SAT = "s"; | ||
| static const String ADV = "r"; | ||
| static const String NOUN = "n"; | ||
| static const String VERB = "v"; | ||
|
|
||
| Lemmatizer::Lemmatizer(const String &path) : path_(path) {} | ||
|
|
||
| Lemmatizer::~Lemmatizer() {} | ||
|
|
||
| Status Lemmatizer::Load() { | ||
| // wninit(path_.c_str()); | ||
|
|
||
| file_map_ = {{ADJ, "adj"}, {ADV, "adv"}, {NOUN, "noun"}, {VERB, "verb"}}; | ||
| pos_numbers_ = {{NOUN, 1}, {VERB, 2}, {ADJ, 3}, {ADV, 4}, {ADJ_SAT, 5}}; | ||
| MORPHOLOGICAL_SUBSTITUTIONS = { | ||
| {NOUN, {{"s", ""}, {"ses", "s"}, {"ves", "f"}, {"xes", "x"}, {"zes", "z"}, {"ches", "ch"}, {"shes", "sh"}, {"men", "man"}, {"ies", "y"}}}, | ||
| {VERB, {{"s", ""}, {"ies", "y"}, {"es", "e"}, {"es", ""}, {"ed", "e"}, {"ed", ""}, {"ing", "e"}, {"ing", ""}}}, | ||
| {ADJ, {{"er", ""}, {"est", ""}, {"er", "e"}, {"est", "e"}}}, | ||
| {ADV, {}}, | ||
| {ADJ_SAT, {{"er", ""}, {"est", ""}, {"er", "e"}, {"est", "e"}}}}; | ||
|
|
||
| pos_names_ = [this] { | ||
| HashMap<int, String> names; | ||
| for (const auto &pair : pos_numbers_) { | ||
| names[pair.second] = pair.first; | ||
| } | ||
| return names; | ||
| }(); | ||
|
|
||
| POS_LIST = {NOUN, VERB, ADJ, ADV}; | ||
|
|
||
| LoadLemmaPosOffsetMap(); | ||
| LoadExceptionMap(); | ||
|
|
||
| return Status::OK(); | ||
| } | ||
|
|
||
| Status Lemmatizer::LoadLemmaPosOffsetMap() { | ||
| for (const auto &pair : file_map_) { | ||
| const String &pos = pair.second; | ||
|
|
||
| std::ifstream file(path_ + "/" + "index." + pos); | ||
| if (!file.is_open()) { | ||
| return Status::InvalidAnalyzerFile(fmt::format("Failed to load wordnet lemmatizer, index.{}", pos)); | ||
| } | ||
|
|
||
| String line; | ||
|
|
||
| while (std::getline(file, line)) { | ||
| if (line.empty() || line[0] == ' ') { | ||
| continue; // Skip comment lines or empty lines | ||
| } | ||
|
|
||
| std::istringstream stream(line); | ||
| try { | ||
| String lemma; | ||
| String pos; | ||
| std::getline(stream, lemma, ' '); | ||
| std::getline(stream, pos, ' '); | ||
|
|
||
| int n_synsets; | ||
| stream >> n_synsets; | ||
| assert(n_synsets > 0); | ||
|
|
||
| int n_pointers; | ||
| stream >> n_pointers; | ||
| // Ignore pointer symbols | ||
| for (int i = 0; i < n_pointers; ++i) { | ||
| String pointer_symbol; | ||
| stream >> pointer_symbol; // Dummy read | ||
| } | ||
|
|
||
| int n_senses; | ||
| stream >> n_senses; | ||
| assert(n_synsets == n_senses); | ||
|
|
||
| // Ignore the number of senses ranked by frequency | ||
| int n_ranked_senses; | ||
| stream >> n_ranked_senses; | ||
|
|
||
| // Get synset offsets | ||
| Vector<int> synset_offsets(n_synsets); | ||
| for (int i = 0; i < n_synsets; ++i) { | ||
| stream >> synset_offsets[i]; | ||
| } | ||
|
|
||
| // Map lemmas and parts of speech to synsets | ||
| lemma_pos_offset_map_[lemma][pos] = synset_offsets; | ||
|
|
||
| if (pos == ADJ) { | ||
| // Duplicate all adjectives indiscriminately | ||
| lemma_pos_offset_map_[lemma][ADJ_SAT] = synset_offsets; | ||
| } | ||
|
|
||
| } catch (const std::exception &) { | ||
| return Status::InvalidAnalyzerFile("Failed to load wordnet lemmatizer"); | ||
| } | ||
| } | ||
| } | ||
| return Status::OK(); | ||
| } | ||
|
|
||
| void Lemmatizer::LoadExceptionMap() { | ||
| for (const auto &pair : file_map_) { | ||
| const auto &pos = pair.first; | ||
| const auto &suffix = pair.second; | ||
|
|
||
| std::ifstream file(path_ + "/" + suffix + ".exc"); | ||
| std::cout << "exception file " << suffix + ".exc" << std::endl; | ||
| exception_map_[pos] = {}; | ||
|
|
||
| String line; | ||
| while (std::getline(file, line)) { | ||
| std::istringstream stream(line); | ||
| String term; | ||
| stream >> term; | ||
| Vector<String> exceptions{std::istream_iterator<String>{stream}, std::istream_iterator<String>{}}; | ||
| exception_map_[pos][term] = exceptions; | ||
| } | ||
| } | ||
| exception_map_[ADJ_SAT] = exception_map_.at(ADJ); | ||
| } | ||
|
|
||
| Vector<String> Lemmatizer::ApplyRules(const Vector<String> &forms, const String &pos) { | ||
| const auto &substitutions = MORPHOLOGICAL_SUBSTITUTIONS.at(pos); | ||
| Vector<String> results; | ||
| for (const auto &form : forms) { | ||
| for (const auto &[old, new_suffix] : substitutions) { | ||
| if (form.size() >= old.size() && form.compare(form.size() - old.size(), old.size(), old) == 0) { | ||
| results.push_back(form.substr(0, form.size() - old.size()) + new_suffix); | ||
| } | ||
| } | ||
| } | ||
| return results; | ||
| } | ||
|
|
||
| Vector<String> Lemmatizer::FilterForms(const Vector<String> &forms, const String &pos) { | ||
| Vector<String> result; | ||
| Set<String> seen; | ||
| for (const auto &form : forms) { | ||
| // Check if form exists in lemma_pos_offset_map_ | ||
| if (lemma_pos_offset_map_.find(form) != lemma_pos_offset_map_.end()) { | ||
| if (lemma_pos_offset_map_[form].find(pos) != lemma_pos_offset_map_[form].end()) { | ||
| if (seen.find(form) == seen.end()) { | ||
| result.push_back(form); | ||
| seen.insert(form); | ||
| } | ||
| } | ||
| } | ||
| } | ||
| return result; | ||
| } | ||
|
|
||
| Vector<String> Lemmatizer::Morphy_(const String &form, const String &pos, bool check_exceptions) { | ||
| const auto &exceptions = exception_map_.at(pos); | ||
|
|
||
| Vector<String> forms; | ||
|
|
||
| if (check_exceptions && exceptions.find(form) != exceptions.end()) { | ||
| forms = exceptions.at(form); | ||
| } else { | ||
| forms = ApplyRules({form}, pos); | ||
| } | ||
|
|
||
| forms.push_back(form); // Include original form | ||
| return FilterForms(forms, pos); | ||
| } | ||
|
|
||
| String Lemmatizer::Lemmatize(const String &form, const String &pos) { | ||
| // If pos is specified, create a vector with that pos, otherwise use all POS_LIST | ||
| Vector<String> parts_of_speech; | ||
| if (!pos.empty()) { | ||
| parts_of_speech.push_back(pos); | ||
| } else { | ||
| parts_of_speech = POS_LIST; | ||
| } | ||
|
|
||
| for (const auto &part : parts_of_speech) { | ||
| auto analyses = Morphy_(form, part); | ||
| if (!analyses.empty()) { | ||
| // Return the first successful analysis | ||
| return analyses[0]; | ||
| } | ||
| } | ||
|
|
||
| return form; | ||
| } | ||
|
|
||
| } // namespace infinity |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,56 @@ | ||
| // Copyright(C) 2024 InfiniFlow, Inc. All rights reserved. | ||
| // | ||
| // Licensed under the Apache License, Version 2.0 (the "License"); | ||
| // you may not use this file except in compliance with the License. | ||
| // You may obtain a copy of the License at | ||
| // | ||
| // | ||
| // https://www.apache.org/licenses/LICENSE-2.0 | ||
| // Unless required by applicable law or agreed to in writing, software | ||
| // distributed under the License is distributed on an "AS IS" BASIS, | ||
| // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| // See the License for the specific language governing permissions and | ||
| // limitations under the License. | ||
|
|
||
| module; | ||
|
|
||
| export module lemmatizer; | ||
|
|
||
| import stl; | ||
| import status; | ||
|
|
||
| namespace infinity { | ||
|
|
||
| export class Lemmatizer { | ||
| public: | ||
| Lemmatizer(const String &path); | ||
|
|
||
| ~Lemmatizer(); | ||
|
|
||
| Status Load(); | ||
|
|
||
| String Lemmatize(const String &form, const String &pos = ""); | ||
|
|
||
| private: | ||
| Status LoadLemmaPosOffsetMap(); | ||
|
|
||
| void LoadExceptionMap(); | ||
|
|
||
| Vector<String> Morphy_(const String &form, const String &pos, bool check_exceptions = true); | ||
|
|
||
| Vector<String> ApplyRules(const Vector<String> &forms, const String &pos); | ||
|
|
||
| Vector<String> FilterForms(const Vector<String> &forms, const String &pos); | ||
|
|
||
| String path_; | ||
|
|
||
| HashMap<String, HashMap<String, Vector<int>>> lemma_pos_offset_map_; | ||
| HashMap<String, HashMap<String, Vector<String>>> exception_map_; | ||
| HashMap<String, int> pos_numbers_; | ||
| HashMap<int, String> pos_names_; | ||
| HashMap<String, String> file_map_; | ||
| HashMap<String, Vector<Pair<String, String>>> MORPHOLOGICAL_SUBSTITUTIONS; | ||
| Vector<String> POS_LIST; | ||
| }; | ||
|
|
||
| } // namespace infinity |
Oops, something went wrong.