-
Notifications
You must be signed in to change notification settings - Fork 27
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #11 from FieldsMedal/hotwords
Add hotwords
- Loading branch information
Showing
11 changed files
with
617 additions
and
23 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
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
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 +1,104 @@ | ||
| // Copyright (c) 2023, 58.com(Wuba) Inc AI Lab. 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 | ||
| // | ||
| // http://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. | ||
| // Modified from DeepSpeech(https://github.com/mozilla/DeepSpeech) | ||
|
|
||
| #include <iostream> | ||
| #include <unordered_map> | ||
|
|
||
| #include "path_trie.h" | ||
| #include "hotwords.h" | ||
| #include "scorer.h" | ||
|
|
||
| HotWordsScorer::HotWordsScorer(const std::unordered_map<std::string, float> &hotwords_dict, const std::vector<std::string>& char_list, | ||
| int window_length, int SPACE_ID, bool is_character_based) { | ||
| this->hotwords_dict = hotwords_dict; | ||
| this->window_length = window_length; | ||
| this->is_character_based = is_character_based; | ||
| this->SPACE_ID = SPACE_ID; | ||
| this->char_list = char_list; | ||
| } | ||
|
|
||
| HotWordsScorer::~HotWordsScorer(){ | ||
| } | ||
|
|
||
| std::string HotWordsScorer::vec2str(const std::vector<int>& input) { | ||
| std::string word; | ||
| for (auto ind : input) { | ||
| word += this->char_list[ind]; | ||
| } | ||
| return word; | ||
| } | ||
|
|
||
| std::pair<int, std::vector<std::string>> HotWordsScorer::make_ngram(PathTrie* prefix){ | ||
| std::vector<std::string> ngram; | ||
| PathTrie* current_node = prefix; | ||
| PathTrie* new_node = nullptr; | ||
| int no_start_token_count = 0; | ||
| for (int order = 0; order < this->window_length; order++) { | ||
| std::vector<int> prefix_vec; | ||
|
|
||
| if (this->is_character_based) { | ||
| new_node = current_node->get_path_vec(prefix_vec, this->SPACE_ID, 1); | ||
| current_node = new_node; | ||
| } else { | ||
| new_node = current_node->get_path_vec(prefix_vec, this->SPACE_ID); | ||
| current_node = new_node->parent; // Skipping spaces | ||
| } | ||
|
|
||
| // reconstruct word | ||
| std::string word = vec2str(prefix_vec); | ||
| ngram.push_back(word); | ||
| no_start_token_count++; | ||
|
|
||
| if (new_node->character == -1) { | ||
| // No more spaces, but still need order | ||
| for (int i = 0; i < this->window_length - order - 1; i++) { | ||
| ngram.push_back(START_TOKEN); | ||
| } | ||
| break; | ||
| } | ||
| } | ||
| std::reverse(ngram.begin(), ngram.end()); | ||
| std::pair<int, std::vector<std::string>> result(this->window_length - no_start_token_count, ngram); | ||
| return result; | ||
| } | ||
|
|
||
| float HotWordsScorer::get_hotwords_score(const std::vector<std::string>& words, int offset) { | ||
| float hotwords_score = 0; | ||
| int words_size = words.size(); | ||
| std::unordered_map<std::string, float>::const_iterator iter; | ||
| for (size_t index = 0; index < words_size; index++) { | ||
| std::string word = ""; | ||
| if (this->is_character_based) { | ||
| // contains at least two chinese characters. | ||
| // words.end()-words.begin()-offset-index+1=words.size()-1-offset-index+1>=2 | ||
| if( words_size - offset - index <= 1) { | ||
| break; | ||
| } | ||
| // chinese characters in fixed window, combining chinese characters into words. | ||
| // word = std::accumulate(words.begin() + offset, words.end() - index, std::string{}); | ||
| word = std::accumulate(words.begin() + offset + index, words.end(), std::string{}); | ||
| } else { | ||
| // word in fixed window, traverse each word in words. | ||
| word = words[index]; | ||
| } | ||
| iter = this->hotwords_dict.find(word); | ||
| if (iter != this->hotwords_dict.end()) { | ||
| hotwords_score += iter->second; | ||
| // break loop after matching the hotwords. | ||
| break; | ||
| } | ||
| } | ||
| return hotwords_score; | ||
| } |
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 +1,47 @@ | ||
| // Copyright (c) 2023, 58.com(Wuba) Inc AI Lab. 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 | ||
| // | ||
| // http://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. | ||
| // Modified from DeepSpeech(https://github.com/mozilla/DeepSpeech) | ||
|
|
||
| #ifndef HOTWORDS_H | ||
| #define HOTWORDS_H | ||
|
|
||
| #include <iostream> | ||
| #include <string> | ||
| #include <unordered_map> | ||
|
|
||
| #include "scorer.h" | ||
|
|
||
| class HotWordsScorer { | ||
| public: | ||
| HotWordsScorer(const std::unordered_map<std::string, float> &hotwords_dict, const std::vector<std::string>& char_list, | ||
| int window_length=4, int SPACE_ID=-1, bool is_character_based=true); | ||
| ~HotWordsScorer(); | ||
|
|
||
| // make ngram for a given prefix | ||
| std::pair<int, std::vector<std::string>> make_ngram(PathTrie *prefix); | ||
|
|
||
| // translate the vector in index to string | ||
| std::string vec2str(const std::vector<int>& input); | ||
|
|
||
| // add hotwords score | ||
| float get_hotwords_score(const std::vector<std::string>& words, int begin_index); | ||
|
|
||
| std::unordered_map<std::string, float> hotwords_dict; | ||
| int window_length; | ||
| int SPACE_ID; | ||
| bool is_character_based; | ||
| std::vector<std::string> char_list; | ||
| }; | ||
|
|
||
| #endif // HOTWORDS_H |
Oops, something went wrong.