Term Weighting Function Research Articles

Time-Weighted Cumulative LSTM Method Using Log Data for Predicting Credit Card Customer Turnover

One of the most important challenges for credit card companies is to retain customers. To achieve this, they predict whether customers will use credit cards and respond preemptively through marketing. Improper prediction leads to an increase in marketing costs; therefore, improving the performance of the prediction model is an important challenge for credit card companies. Therefore, credit card companies are working toward increasing the predictive power of the model, but the data used for learning are limited to formal data, such as card payment data; the log history stored on the application is not used. Prediction models using sequence data, such as log history, are being studied. However, no studies so far have appropriately utilized card payment data, log history, and information on the timing of the elements constituting the sequence. In this paper, we propose a method for predicting credit card turnover using application log and credit card payment data. The method embeds log data and weights each log using a term-weighting function. In addition, the time difference between the day customers visit the app and the last day of the month in which they leave is calculated. The weight value is higher when the timespan is shorter. The weighted log data were trained with a 2-layer LSTM model and its output layer was combined with the output layer trained with the card payment data for customer turnover prediction. It was verified that the model presented in this paper had a higher performance than the model using only card payment data and the model using only a time-weighted log sequence.

Learning term weights by overfitting pairwise ranking loss

A search engine strikes a balance between effectiveness and efficiency to retrieve the best documents in a scalable way. Recent deep learning-based ranker methods are proving to be effective and improving the state-of-the-art in relevancy metrics. However, as opposed to index-based retrieval methods, neural rankers like bidirectional encoder representations from transformers (BERT) do not scale to large datasets. In this article, we propose a query term weighting method that can be used with a standard inverted index without modifying it. Query term weights are learned using relevant and irrelevant document pairs for each query, using a pairwise ranking loss. The learned weights prove to be more effective than term recall which is a probabilistic relevance feedback, previously used for the task. We further show that these weights can be predicted with a BERT regression model and improve the performance of both a BM25 based index and an index already optimized with a term weighting function.

Topic Models Ensembles for AD-HOC Information Retrieval

Ad hoc information retrieval (ad hoc IR) is a challenging task consisting of ranking text documents for bag-of-words (BOW) queries. Classic approaches based on query and document text vectors use term-weighting functions to rank the documents. Some of these methods’ limitations consist of their inability to work with polysemic concepts. In addition, these methods introduce fake orthogonalities between semantically related words. To address these limitations, model-based IR approaches based on topics have been explored. Specifically, topic models based on Latent Dirichlet Allocation (LDA) allow building representations of text documents in the latent space of topics, the better modeling of polysemy and avoiding the generation of orthogonal representations between related terms. We extend LDA-based IR strategies using different ensemble strategies. Model selection obeys the ensemble learning paradigm, for which we test two successful approaches widely used in supervised learning. We study Boosting and Bagging techniques for topic models, using each model as a weak IR expert. Then, we merge the ranking lists obtained from each model using a simple but effective top-k list fusion approach. We show that our proposal strengthens the results in precision and recall, outperforming classic IR models and strong baselines based on topic models.

A Term Weighted Neural Language Model and Stacked Bidirectional LSTM Based Framework for Sarcasm Identification

Sarcasm identification on text documents is one of the most challenging tasks in natural language processing (NLP), has become an essential research direction, due to its prevalence on social media data. The purpose of our research is to present an effective sarcasm identification framework on social media data by pursuing the paradigms of neural language models and deep neural networks. To represent text documents, we introduce inverse gravity moment based term weighted word embedding model with trigrams. In this way, critical words/terms have higher values by keeping the word-ordering information. In our model, we present a three-layer stacked bidirectional long short-term memory architecture to identify sarcastic text documents. For the evaluation task, the presented framework has been evaluated on three-sarcasm identification corpus. In the empirical analysis, three neural language models (i.e., word2vec, fastText and GloVe), two unsupervised term weighting functions (i.e., term-frequency, and TF-IDF) and eight supervised term weighting functions (i.e., odds ratio, relevance frequency, balanced distributional concentration, inverse question frequency-question frequency-inverse category frequency, short text weighting, inverse gravity moment, regularized entropy and inverse false negative-true positive-inverse category frequency) have been evaluated. For sarcasm identification task, the presented model yields promising results with a classification accuracy of 95.30%.

Learning to Weight for Text Classification

In information retrieval (IR) and related tasks, term weighting approaches typically consider the frequency of the term in the document and in the collection in order to compute a score reflecting the importance of the term for the document. In tasks characterized by the presence of training data (such as text classification) it seems logical that the term weighting function should take into account the distribution (as estimated from training data) of the term across the classes of interest. Although “supervised term weighting” approaches that use this intuition have been described before, they have failed to show consistent improvements. In this article, we analyze the possible reasons for this failure, and call consolidated assumptions into question. Following this criticism, we propose a novel supervised term weighting approach that, instead of relying on any predefined formula, learns a term weighting function optimized on the training set of interest; we dub this approach Learning to Weight (LTW). The experiments that we run on several well-known benchmarks, and using different learning methods, show that our method outperforms previous term weighting approaches in text classification.

Organizing photographs with geospatial and image semantics

The popularity of GPS-equipped gadgets and mapping mashup applications has motivated the growth of geotagged Web resources as well as georeferenced multimedia applications. More and more research attention have been put on mining collaborative knowledge from mass user-contributed geotagged contents. However, little attention has been paid to generating high-quality geographical clusters, which is an important preliminary data-cleaning process for most geographical mining works. Previous works mainly use geotags to derive geographical clusters. Simply using one channel information is not sufficient for generating distinguishable clusters, especially when the location ambiguity problem occurs. In this paper, we propose a two-level clustering framework to utilize both the spatial and the semantic features of photographs for clustering. For the first-level geoclustering phase, we cluster geotagged photographs according to their spatial ties to roughly partition the dataset in an efficient way. Then we leverage the textual semantics in photographs' annotation to further refine the grouping results in the second-level semantic clustering phase. To effectively measure the semantic correlation between photographs, a semantic enhancement method as well as a new term weighting function have been proposed. We also propose a method for automatic parameter determination for the second-level spectral clustering process. Evaluation of our implementation on real georeferenced photograph dataset shows that our algorithm performs well, producing distinguishable geographical cluster with high accuracy and mutual information.

BM25t: a BM25 extension for focused information retrieval

This paper addresses the integration of XML tags into a term-weighting function for focused XML information retrieval (IR). Our model allows us to consider a certain kind of structural information: tags that represent a logical structure (e.g., title, section, paragraph, etc.) as well as other tags (e.g., bold, italic, center, etc.). We take into account the influence of a tag by estimating the probability for this tag to distinguish relevant terms from the others. Then, these weights are integrated in a term-weighting function. Experiments on a large collection from the INEX 2008 XML IR evaluation campaign showed improvements on focused XML retrieval.

Experiments in term weighting for novelty mining

Obtaining new information in a short time is becoming crucial in today’s economy. A lot of information both offline or online is easily acquired, exacerbating the problem of information overload. Novelty mining detects documents/sentences that contain novel or new information and presents those results directly to users ( Tang, Tsai, & Chen, 2010). Many methods and algorithms for novelty mining have previously been studied, but none have compared and discussed the impact of term weighting on the evaluation measures. This paper performed experiments to recommend the best term weighting function for both document and sentence-level novelty mining.

A ranking algorithm for query expansion based on the term's appearing probability in the single document

PurposeThe purpose of this paper is to establish a new approach for solving the expansion term problem.Design/methodology/approachThis study develops an expansion term weighting function derived from the valuable concepts used by previous approaches. These concepts include probability measurement, adjustment according to situations, and summation of weights. Formal tests have been conducted to compare the proposed weighting function with the baseline ranking model and other weighting functions.FindingsThe results reveal stable performance by the proposed expansion term weighting function. It proves more effective than the baseline ranking model and outperforms other weighting functions.Research limitations/implicationsThe paper finds that testing additional data sets and potential applications to real working situations is required before the generalisability and superiority of the proposed expansion term weighting function can be asserted.Originality/valueStable performance and an acceptable level of effectiveness for the proposed expansion term weighting function indicate the potential for further study and development of this approach. This would add to the current methods studied by the information retrieval community for culling information from documents.

Probabilistic static pruning of inverted files

Information retrieval (IR) systems typically compress their indexes in order to increase their efficiency. Static pruning is a form of lossy data compression: it removes from the index, data that is estimated to be the least important to retrieval performance, according to some criterion. Generally, pruning criteria are derived from term weighting functions, which assign weights to terms according to their contribution to a document's contents. Usually, document-term occurrences that are assigned a low weight are ruled out from the index. The main assumption is that those entries contribute little to the document content. We present a novel pruning technique that is based on a probabilistic model of IR. We employ the Probability Ranking Principle as a decision criterion over which posting list entries are to be pruned. The proposed approach requires the estimation of three probabilities, combining them in such a way that we gather all the necessary information to apply the aforementioned criterion. We evaluate our proposed pruning technique on five TREC collections and various retrieval tasks, and show that in almost every situation it outperforms the state of the art in index pruning. The main contribution of this work is proposing a pruning technique that stems directly from the same source as probabilistic retrieval models, and hence is independent of the final model used for retrieval.

Enlightening the sun

This article presents an approach to browse collections of web pages about music artists by means of descriptive terms and multimedia content. To this end, a user interface called Three-Dimensional Co-Occurrence Browser (3D-COB) is introduced. 3D-COB automatically extracts and weights terms from artist-related web pages. This textual information is complemented with information on the multimedia content found on the web pages. For the user interface of 3D-COB, we elaborated a three-dimensional extension of the Sunburst visualization technique. The hierarchical data to be visualized is obtained by analyzing the web pages for combinations of co-occurring terms that are highly ranked by a term weighting function. We further investigated, in a first user study, different term weighting strategies to generate the visualization. A second user study was carried out to assess ergonomic aspects of 3D-COB, especially its usefulness for gaining a quick overview of a set of web pages and for efficiently browsing within this set.

An axiomatic comparison of learned term-weighting schemes in information retrieval: clarifications and extensions

Machine learning approaches to information retrieval are becoming increasingly widespread. In this paper, we present term-weighting functions reported in the literature that were developed by four separate approaches using genetic programming. Recently, a number of axioms (constraints), from which all good term-weighting schemes should be deduced, have been developed and shown to be theoretically and empirically sound. We introduce a new axiom and empirically validate it by modifying the standard BM25 scheme. Furthermore, we analyse the BM25 scheme and the four learned schemes presented to determine if the schemes are consistent with the axioms. We find that one learned term-weighting approach is consistent with more axioms than any of the other schemes. An empirical evaluation of the schemes on various test collections and query lengths shows that the scheme that is consistent with more of the axioms outperforms the other schemes.

Laboratory experiments with Okapi: participation in the TREC programme

A brief review of the history of laboratory testing of information retrieval systems focuses on the idea of a general‐purpose test collection of documents, queries and relevance judgements. The TREC programme is introduced in this context, and an overview is given of the methods used in TREC. The Okapi team’s participation in TREC is then discussed. The team has made use of TREC to improve some of the automatic techniques used in Okapi, specifically the term weighting function and the algorithms for term selection for query expansion. The consequence of this process has been a very good showing for Okapi in terms of the TREC evaluation results. Some of the issues around the much more difficult problem of interactive evaluation in TREC are then discussed. Although some interesting interactive experiments have been performed at TREC, the problems of reconciling the requirements of the laboratory context with the concerns of interactive retrieval are still largely unresolved.

Comparing and combining the effectiveness of latent semantic indexing and the ordinary vector space model for information retrieval

A retrieval system was built to find individuals with appropriate expertise within a large research establishment on the basis of their authored documents. The expert-locating system uses a new method for automatic indexing and retrieval based on singular value decomposition, a matrix decomposition technique related to factor analysis. Organizational groups, represented by the documents they write, and the terms contained in these documents, are fit simultaneously into a 100-dimensional “semantic” space. User queries are positioned in the semantic space, and the most similar groups are returned to the user. Here we compared the standard vector-space model with this new technique and found that combining the two methods improved performance over either alone. We also examined the effects of various experimental variables on the system's retrieval accuracy. In particular, the effects of: term weighting functions in the semantic space construction and in query construction, suffix stripping, and using lexical units larger than a single word were studied.

Use of dynamic discrimination values in a document retrieval system

The use of discrimination values as a term weighting function in document retrieval systems is examined. It is shown that regular discrimination values are too costly to compute after every update to the data base. Dynamic discrimination values that are easy to update are defined for use as approximations to regular values. Experiments are performed comparing regular vs. dynamic discrimination values. Actual user queries from an operational data base are used to evaluate dynamic discrimination values in a production environment. Generalized forms of normalized recall and precision are used as evaluation measures. Retrieval results indicate statistically significant improvements using dynamic discrimination weighting.