added RetrieveSentences.py (#73)

2026-09-02 12:01:12 +08:00 · 2017-10-28 15:05:52 -04:00
parent 491f0b32d5
commit 00ef2d0fde
2 changed files with 115 additions and 0 deletions
@@ -0,0 +1,46 @@
+## Retrieve Sentences
+
+#### 1. Clone [Anserini](https://github.com/castorini/Anserini.git) and [Castor](https://github.com/castorini/Castor.git)
+```bash
+git clone https://github.com/castorini/Anserini.git
+git clone https://github.com/castorini/Castor.git
+```
+
+Your directory structure should look like
+```
+├── Anserini
+└── Castor
+```
+
+#### 2. Compile Anserini
+
+```bash
+cd Anserini
+mvn package
+``` 
+
+This creates `anserini-0.0.1-SNAPSHOT.jar` at `Anserini/target`
+
+#### 3. Download Dependencies
+- Download the index from [here](https://drive.google.com/open?id=0B2u_nClt6NbzcllGZTVtX2Q4bkk)
+- Download the Google word2vec file from [here](https://drive.google.com/drive/folders/0B2u_nClt6NbzNWJkWExmaklYNTA?usp=sharing)
+
+#### 4. Run the following command
+
+```bash
+python ./anserini_dependency/RetrieveSentences.py
+```
+
+Possible parameters are: 
+
+| option         | input format | default | description |
+|----------------|--------------|---------|-------------|
+| `-index`   | string      | N/A     | Path of the Lucene index            |
+| `-embeddings`   | string       | ""     | Path of the word2vec index            |
+| `-topics`   | string       | ""     | topics file            |
+| `-query`   | string       | ""     | a single query            |
+| `-hits`   | [1, inf)       | 100     | max number of hits to return            |
+| `-scorer`   | string       | Idf     | passage scores (Idf or Wmd)            |
+| `-k`   | [1, inf)       | 1     | top-k passages to be retrieved            |
+  
+Note: Either a query or a topic must be passed in as an argument; they can't be both empty.
@@ -0,0 +1,69 @@
+import argparse
+
+import jnius_config
+jnius_config.set_classpath("../Anserini/target/anserini-0.0.1-SNAPSHOT.jar")
+from jnius import autoclass
+
+
+class CallRetrieveSentences:
+    """Python class built to call RetrieveSentences
+    Attributes
+    ----------
+    rs : io.anserini.qa.RetrieveSentences
+        The RetrieveSentences object
+    args : io.anserini.qa.RetrieveSentences$Args
+        The arguments for constructing RetrieveSentences object as well as calling getRankedPassages
+
+    """
+
+    def __init__(self, args):
+        """
+        Constructor for the CallRetrieveSentences class.
+
+        Parameters
+        ----------
+        args : argparse.Namespace
+            Arguments needed for constructing an instance of RetrieveSentences class
+        """
+        RetrieveSentences = autoclass("io.anserini.qa.RetrieveSentences")
+        Args = autoclass("io.anserini.qa.RetrieveSentences$Args")
+        String = autoclass("java.lang.String")
+
+        self.args = Args()
+        index = String(args.index)
+        self.args.index = index
+        embeddings = String(args.embeddings)
+        self.args.embeddings = embeddings
+        topics = String(args.topics)
+        self.args.topics = topics
+        query = String(args.query)
+        self.args.query = query
+        self.args.hits = int(args.hits)
+        scorer = String(args.scorer)
+        self.args.scorer = scorer
+        self.args.k = int(args.k)
+        self.rs = RetrieveSentences(self.args)
+
+    def getRankedPassages(self):
+        """
+        Call RetrieveSentneces.getRankedPassages
+        """
+        self.rs.getRankedPassages(self.args)
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description='Retrieve Sentences')
+    parser.add_argument("-index", help="Lucene index", required=True)
+    parser.add_argument("-embeddings", help="Path of the word2vec index", default="")
+    parser.add_argument("-topics", help="topics file", default="")
+    parser.add_argument("-query", help="a single query", default="")
+    parser.add_argument("-hits", help="max number of hits to return", default=100)
+    parser.add_argument("-scorer", help="passage scores", default="Idf")
+    parser.add_argument("-k", help="top-k passages to be retrieved", default=1)
+
+    args_raw = parser.parse_args()
+    rs = CallRetrieveSentences(args_raw)
+    rs.getRankedPassages()
+
+
+
+