Combining the language model and inference network approaches to retrieval

By Donald Metzler, W. Bruce Croft. 2004

322 citations recorded by the source corpus.

The inference network retrieval model, as implemented in the InQuery search engine, allows for richly structured queries. However, it incorporates a form of ad hoc tf.idf estimates for word probabilities. Language modeling offers more formal estimation techniques. In this paper we combine the language modeling and inference network approaches into a single framework. The resulting model allows structured queries to be evaluated using language modeling estimates. We explore the issues involved, such as combining beliefs and smoothing of proximity nodes. Experimental results are presented comparing the query likelihood model, the InQuery system, and our new model. The results reaffirm that high quality structured queries outperform unstructured queries and show that our system consistently achieves higher average precision than InQuery.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Information Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Information Systems

Citations in the corpus

Anserini (2018, background): “An incomplete list includes Lemur/Indri [38, 39], Galago [15], Terrier [35, 42], ATIRE [51], Ivory [30], JASS [31], MG4J [12], [60, 61].”
Indri at TREC 2004: Terabyte Track (2004, methodology): “The retrieval model implemented in the Indri search engine is an enhanced version of the model described in [9], which combines the language modeling [12, 3] and inference network [13] approaches to information retrieval.”
An Exploration of Approaches to Integrating Neural Reranking Models in Multi-Stage Ranking Architectures (2017, background): “To address this challenge, many academic research groups have built and shared open-source search engines with the broader community—prominent examples include Lemur/Indri [9, 10] and Terrier [7, 13].”
Find-similar: similarity browsing as a search tool (2006, methodology): “We used both the language modeling approach to information retrieval [25] and its combination with the inference network approach [21] as implemented in the Lemur [18] and Indri [33] retrieval systems.”
A Survey of Answer Extraction Techniques in Factoid Question Answering (2006, methodology): “Two of the most popular search engine used by many QA systems are Indri (Metzler and Croft 2004) and Lucene 1 .”
Toward Reproducible Baselines: The Open-Source IR Reproducibility Challenge (2016, methodology): “In total, developers from seven open-source systems participated (in alphabetical order): ATIRE [16], Galago [6], Indri [10,12], JASS [9], Lucene [2], MG4J [3], and Terrier [14].”
Learning to Reweight Terms with Distributed Representations (2015, methodology): “True term recall weight or term recall weight estimates can used by a language model in similar fashion as [9], as shwon in Eq.”
The Lucene for Information Access and Retrieval Research (LIARR) Workshop at SIGIR 2017 (2017, methodology): “An incomplete list includes SMART [18], Lemur/Indri [15, 16], Galago [8], Terrier [14, 17], ATIRE [20], Ivory [12], JASS [13], MG4J [6], Wumpus [21], and Zeair [25].”
Similarity of business process models: Metrics and evaluation (2011, methodology): “2 We also used a traditional search engine (the Indri search engine [15]) to answer these same queries.”
Old dogs are great at new tricks: column stores for ir prototyping (2014, background): “Today, these techniques can be quite complex, especially with “structured queries” that span multiple nested clauses with a panoply of query operators [11].”